Available via license: CC BY-NC 4.0
Content may be subject to copyright.
JOURNAL OF MECHANICAL ENGINEERING AND SCIENCES
ISSN: 2289-4659 e-ISSN: 2231-8380
VOLUME 18, ISSUE 4, 2024, 10350 - 10381
DOI: https://doi.org/10.15282/jmes.18.4.2024.9.0815
*CORRESPONDING AUTHOR | Y. C. Tan | tanyongchai@segi.edu.my
© 2024 The Auth or(s). Published by Universiti Malaysia Pahang Al-Sul tan Abdullah Publishin g. This is an open access article under the CC BY -NC 4.0 license 10350
REVIEW ARTICLE
Indoor positioning system for warehouse environment: A scoping review
X. D. Zhang1,2, Y. C. Tan2*, V. C. Tai2, Y. N. Hao3
1 Department of Automation Engineering, Henan Polytechnic Institute, 473000, Nanyang City, Henan Province, China
2 Centre for Sustainable Design, Modelling and Simulation, Faculty of Engineering, Built Environment, and Information Technology, SEGi University,
47810 Petaling Jaya, Selangor, Malaysia
Phone: +603 61451777; Fax.: +60361452725
3 Department of Electrical Engineering, Taiyuan Institute of Technology. No.31 Xinlan Road, Taiyuan, Shanxi 030008, China
ABSTRACT - Advanced technologies and automation, driven by Indoor Positioning Systems
(IPS), transform businesses by enhancing efficiency, intelligence, and digitalization. Despite
the critical role of IPS, there remains a lack of comprehensive reviews focusing specifically on
their applications in warehouse inventory management. To bridge this gap and provide
actionable insights for both research and practical implementation, this study conducts a
systematic literature review following the PRISMA checklist. Centered around three key
research questions, this review explores
the scope of IPS applications in warehouse
environments,
the specific technologies employed, and the methods to evaluate IPS
performance. This paper analyzes the fundamental principles and recent applications of
widely adopted indoor positioning technologies, including Wi-Fi, UWB, RFID, VLC, IMU,
Computer Vision, and LiDAR. Furthermore, this paper evaluates IPS technologies through five
key evaluation criteria, highlighting their advantages, limitations, and challenges. This study
provides a comprehensive understanding of IPS technologies in warehouse inventory
management, offering actionable methods to evaluate their performance. The insights
presented aim to deliver strong decision support for researchers and practitioners seeking to
optimize inventory operations in warehouse environments.
ARTICLE HISTORY
Received
:
02nd Dec. 2023
Revised
:
18th Dec. 2024
Accepted
:
20th Dec. 2024
Published
:
30th Dec. 2024
KEYWORDS
Indoor positioning system
Indoor localization
Positioning technology
Warehouse environment
Inventory management
Environmental sustainability
1. INTRODUCTION
In recent years, indoor mobile robots have increasingly changed our lives thanks to the rapid development of robotics
and sensor technologies. Boston Dynamics' Spot and Atlas robots are already proficient at assisting humans with complex
tasks in everyday scenarios [1]. Additionally, an increasing number of robotic products are being used in the fields of
industrial automation [2], warehousing and logistics [3], surveying and mapping [4], medical care [5], disaster response,
and home services [6] to replace humans in performing repetitive and laborious tasks and to reduce human errors. As
robotic technology continues to evolve, a greater diversity of fields can be expected to benefit from this technology in the
future. Due to indoor environments' intricate and dynamic properties, Indoor Positioning Systems (IPS) have emerged as
a significant gap in the advancement of location-based technologies. Unlike most outdoor positioning systems, which rely
on the Global Navigation Satellite System (GNSS), IPS has challenges because GNSS signals are severely attenuated in
indoor environments [7]. This attenuation can significantly reduce the accuracy of positioning technologies, such as
satellite-based navigation systems, when used indoors, posing considerable challenges to developing effective indoor
positioning systems [8]. Therefore, the research of IPS is attracting significant attention.
IPS has garnered considerable attention and is useful in various indoor scenarios because most indoor mobile
equipment operations rely on accurate positioning information. IPS can provide precise positioning information within
indoor environments by utilizing a variety of sensors, wireless communications, and advanced positioning algorithms [9].
Sensors such as Wi-Fi, Bluetooth, Radio Frequency Identification Device (RFID), Ultra-Wideband (UWB), ultrasound,
infrared, vision sensors, Light Detection and Ranging (LiDAR), and inertial measurement units (IMU) are leveraged by
these systems to capture and analyze data related to signal strength, time delays, distances, directions, and angles. These
systems offer positioning and tracking information in indoor environments, facilitating various applications, including
indoor navigation. With the development of smart warehouses, mobile intelligent equipment is beginning to assist humans
in warehouse operations such as inventory review, internal logistics, cycle counting, and stocktaking [9]. IPS plays a
pivotal role in the advancement of modern smart warehouses. The basic characteristics of smart warehouses can be
classified into the following categories: information interconnection, equipment automation, process integration, and
environmental sustainability [10]. IPS is beneficial in all these aspects, especially for mobile equipment in warehouses
represented by autonomous mobile robots [11]. For example, when IPS is applied to smart forklifts or Automated Guided
Vehicles (AGVs) in a smart warehouse system [12], it allows these devices to run smoothly, and their movements are not
restricted by changes in logistics activities or shelf layouts [13]. This technology also ensures real-time goods counting
in dynamically changing warehouse environments, enabling faster goods picking, increased accuracy in warehouse
operations, reduced operational costs, and minimized manual errors [14]. Combining robots, drones, and self-driving cars
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10351
can achieve partial or complete autonomy for most warehouse operations tasks [15]. These systems are instrumental in
improving operational efficiency across various domains, including inventory management, logistics, and warehouse
digitization. Among these applications, inventory management is a crucial component of warehouse operations, directly
influencing stock reliability, economic performance, and productivity.
However, to the author's best knowledge, no relevant scoping reviews have been conducted regarding IPS applications
to warehouses. The main contributions of this review are:
i) Describing current applications of IPS in warehouse environments.
ii) Offering a comprehensive review of the current technology path for IPS applied in inventory-related tasks.
iii) Proposing a framework for evaluating different IPS technologies applied in inventory management.
iv) Providing guidance and prospects for future research on IPS technologies in inventory management, highlighting
unresolved challenges and potential innovations.
The article is structured as follows: Section 1 outlines the background of IPS applications in warehouse environments,
explains the significance and necessity of focusing on inventory management, and clarifies the objectives of this scoping
review. Section 2 presents the research methodology used for this scoping review. Section 3 presents relevant results
related to the research objectives, focusing on inventory management applications. Section 4 discusses the findings,
proposes practical recommendations, and concludes the review.
2. SCOPING REVIEW METHODOLOGY
The methodology of this research adheres to the scoping methodological framework proposed by Arksey and Hilary
[16] and the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews
(PRISMA-ScR) execution standards set forth by Tricco and O'Malley [17]. Additionally, techniques from other review
research methods, such as systematic review and meta-analyses, are referenced [18]. The ScR research method allows for
a comprehensive and objective review of research in a specific field. Utilizing the standardized Scoping Reviews
(PRISMA-ScR) Checklist enhances the efficiency of the review process, making the results valuable to readers,
policymakers, and practitioners.
Figure 1. Scoping review process [16]
The PRISMA-ScR Checklist [17] and other review studies that employed this method [20-22] were synthesised. The
research methodology primarily encompasses stages such as "Clear Definition and Searching Preparation," "Searching
and Selection of Data," and "Data Processing," which can be further subdivided into seven detailed steps. The process of
the scoping review undertaken in this paper is depicted in Figure 1.
2.1 STAGE 1: Definition and Searching Preparation
2.1.1 The key terms identification
The criteria of the PRISMA-ScR execution standard [12] are followed in this section for a pre-search investigation
preparation process. Initially, the focus is on identifying key terms and providing accurate conceptual explanations for
subsequent searches. However, it was found that the descriptions or concepts of these terms often lack consistency. Key
concepts, similar terms, and near-synonyms related to the main research questions RQ1, RQ2, and RQ3 are synthesized
and presented in Table 1, accompanied by brief explanations.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10352
Table 1. Key terms
Key terms
Abbreviation &
Synonyms Related terms Explanation
Warehouse
None
Cargo, Inventory, Shelf, Stock,
A warehouse is a large facility for
storing goods, including storage space,
inventory management, and loading
areas.
Indoor
Positioning
IPS
indoor location,
indoor navigation,
Odometry
Sensors, Signal Sources,
Mapping and Localization,
Indoor positioning refers to the
technology or methods used to track the
location of objects or individuals within
indoor environments.
Mobile
Robot
Mobile Robotic,
UAV (Unmanned Aerial Vehicle),
MAV (Micro Aerial Vehicle)
AGV, Drone,
In a warehouse setting, mobile robots,
often called logistics or storage robots,
are automated devices that handle
inventory and logistics tasks.
Algorithm,
Technology
Techniques,
Method
Application
The technology or algorithm mentioned
here pertains to the key technologies
used in the IPS for mobile robots in a
warehouse setting.
The initial search was conducted to refine the scoping review protocol, enhance the research questions, and adjust the
search terms, upon which a more precise formal search would be conducted. The key terms employed in the search are
defined in Table 1, and their synonymous expressions are elaborated.
2.1.2 Identification of research question
A comprehensive scoping review [17] addressing the questions outlined in Table 2 is aimed to be provided by this
study. Through this approach, an in-depth exploration of historical development, a thorough comparison of diverse IPS
technologies suitable for indoor warehouse environments, and the identification of research trends specifically applicable
to inventory management within warehouse settings are offered to the reader.
Table 2. Ultimate tensile strength values and elongation to fracture
No.
Research Questions
Goal
RQ1
What are the current IPS applications in a
warehouse environment?
Explore the applications of IPS in a warehouse environment,
specifically focusing on identifying the role and criticality of IPS
implementation in inventory management.
RQ2
What current IPS technologies are utilised
explicitly
for inventory management in
warehouse environments?
Identify the techniques applied to IPS in a warehouse
environment for inventory management, focusing on the key
methods used for implementation and evaluating the advantages
and limitations of each technology in enhancing inventory
tracking and control.
RQ3
How to evaluate different IPS inventories
for logistics management applications in
a warehouse environment?
Develop a comprehensive evaluation framework for comparing
IPS technologies in warehouse environments, focusing on their
effectiveness, efficiency, and suitability for inventory
management.
2.1.3 Determine the information source
Based on the findings presented in Table 3, it can be concluded that the relevant search terms for the scope of this
scoping review include warehouse, indoor positioning, robot, navigation, and other related terms. Web of Science (WOS),
Scopus, and Institute of Electrical and Electronics Engineers (IEEE) were selected as the search databases because nearly
all peer-reviewed literature in the field of engineering technology application can be searched through the combination
of these three databases.
2.2 STAGE 2: Data Searching and Selection
The flow chart of article inclusion and exclusion, based on the PRISMA execution standard [18], is displayed in
Figure 2. The steps for search preparation, sorting of search results, screening abstracts and conclusions, and full-text
screening are included. Steps IV and Ⅴ will be utilized to detail how articles for inclusion in the review scope were
obtained.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10353
Figure 2. Flow chart of article inclusion and exclusion based on PRISMA criteria [18]
2.2.1 Searching limitations and keywords
The search strategy was designed to ensure comprehensive coverage of relevant literature on IPS technologies in
warehouse environments. Searches were conducted in three major academic databases, Web of Science (WOS), IEEE
Xplore, and Scopus, selected for their extensive engineering and technology-related publications indexing. The search
queries combined keywords related to indoor positioning systems (e.g., "Indoor Positioning System," "IPS," "indoor
localization") with terms specific to warehouse applications (e.g., "warehouse," "inventory management," "logistics").
Boolean operators such as "AND," "OR," and "NOT" were employed to refine the search scope, as illustrated in Table 3.
Table 3. Searching keywords and limitations
Database
Search terms
Scopus
(TITLE-ABS-KEY ( warehouse ) AND TITLE-ABS-
KEY ( "robot*" OR "drone" OR "UAV" OR "AGV" OR "MAV" ) AND TITLE-ABS-
KEY ( "indoor positioning" OR "indoor
localization" OR "navigation" ) ) AND PUBYEAR > 2013 AND PUBYEAR < 2024
AND ( LIMIT-TO ( DOCTYPE , "ar" ) OR LIMIT-TO ( DOCTYPE , "re" ) ) AND ( LIMIT-
TO ( SUBJAREA , "ENGI" ) OR LIMIT-TO ( SUBJAREA , "COMP" ) )
WOS
(TS=(warehouse)) AND (TS=(indoor positioning) OR TS=(indoor localization) OR
TS=(navigation)) AND (TS=(robot*) OR TS =(drone) OR TS =(UAV) OR TS =(AGV) OR TS
=( MAV))
Refined By: Publication Years: 2014 2015 or 2023 or 2022 or 2021 or 2020 or 2019 or 2018 or
2017 or 2016 or 2015 or 2014, Document Types: Article
IEEE
("Abstract”: warehouse) AND ("Full Text & Metadata":"robot*" OR "Full Text &
Metadata":"drone" OR "Full Text & Metadata":"AGV" OR "Full Text & Metadata":"MAV" OR
"Full Text & Metadata":"UAV") AND ("Abstract":"indoor positioning" OR "Abstract":"indoor
localization" OR "Abstract":"navigation")
Filters Applied: Journals2014 - 2023
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10354
Specific filtering criteria were applied during the search to improve relevance and reliability. These criteria included:
Articles published in English, Peer-reviewed journal articles and conference papers, Publications from the past decade
(2014–2023); and a search scope limited to the title, abstract, and keyword sections. An initial search was conducted in
June 2023 using these criteria, resulting in 100 articles from Scopus, 101 articles from WOS, and 17 from IEEE. After
merging the results from all three databases and eliminating duplicates, 112 articles were retained for further filtering.
The search process followed the PRISMA flowchart depicted in Figure 1. As the initial search results were reviewed, the
search strategy was iteratively refined by adjusting keywords and Boolean operators to ensure comprehensive coverage.
This iterative refinement maximized the inclusion of relevant studies while maintaining a focus on IPS technologies
applied to warehouse environments.
2.2.2 Screening and exclusion
The inclusion and exclusion criteria were applied to ensure the selected articles aligned with the study's objectives.
The exclusion criteria are presented in Table 4, emphasizing the selection of studies directly relevant to IPS applications
in inventory management within warehouse environments. Articles were included if they Discussed IPS technologies
specifically in the context of warehouse environments. Focused on inventory management, logistics, or warehouse
digitization tasks. Provided experimental data, theoretical analysis, or practical case studies relevant to IPS applications.
Articles were excluded if they focused solely on outdoor positioning systems or technologies unrelated to warehouse
operations. Lacked detailed descriptions of IPS implementation or applications and were review papers or non-peer-
reviewed publications. The screening process was conducted in two stages: titles and abstracts were reviewed to eliminate
irrelevant articles, followed by a full-text review to ensure alignment with the inclusion criteria. In this phase, the 112
selected articles were reviewed thoroughly. Focus was placed on their titles, abstracts, and keywords to ascertain
alignment with the research topic. After the abstracts and conclusions were screened, 64 articles were excluded for not
matching the subject of this study based on conditions EC1, EC2, and EC3. A comprehensive review of the remaining 48
full-text articles was then conducted. It was identified that 10 of these articles neither primarily focused on IPS nor
provided detailed information about the implementation process of IPS. As a result, these 10 articles were excluded from
further analysis. The methodology employed to extract and analyze data from the remaining 38 selected articles will be
elaborated upon in the subsequent sections.
Table 4. Exclusion criteria
No.
Exclusion criteria
Description
EC1
Exclusion of research that does not
involve IPS
Literature not primarily focused on IPS is excluded, as it is deemed
irrelevant to the study
EC2
Exclusion of items not relevant to
inventory management in a warehouse
environment
Literature in which IPS is not applied, or is not intended to be
applied, in the warehouse environment is excluded
EC3
Exclusion the literature that without real
experiments, the IPS
Literature that has not been tested in practice or whose simulation
tests do not resemble real-
world warehouse environments is
excluded
2.3 STAGE 3: Data Processing
2.3.1 Data Extraction
The literature information included in the scoping review was categorized into three types: foundational data,
distinctive data, and experimental data. This categorization facilitated the organization and comprehensive analysis of the
extracted information. Foundational data included the title, author, publication year, keywords, and journal. A clear
timeline of technological development in this field was outlined through the analysis of foundational data. Development
trends were further analyzed by incorporating distinctive data. Distinctive data comprised research (application)
objectives, conclusions, and future work. A clear understanding of the research problem's focal points, urgent issues
requiring attention, and current challenges was gained through the organization and collection of distinctive data from
each article. Experimental data included technologies used in the study, experimental (testing) platforms, experimental
(testing) environments, accuracy, cost, energy efficiency, and scalability. A comprehensive analysis of the strengths and
weaknesses of various techniques, methods, and algorithms was conducted by organising detailed experimental data and
formulating a comprehensive evaluation framework. A standardized form developed in Microsoft Excel was employed
to uniformly record data from articles that had passed the initial filtering and selection steps to ensure a systematic data
extraction process.
2.3.2 Data Analysis
Quantitative and qualitative analysis was applied to the data and information extracted from the selected articles. This
step involved the organization and analysis of data for RQ1, RQ2, and RQ3. To comprehensively address RQ1, the basic
information of the included articles was first compiled. Based on the objectives of each study, the direction of IPS
applications in warehouse environments over the past decade and the proportion of relevant literature were analyzed. For
RQ2, both foundational and experimental data were utilized to generate a timeline that captures the development of indoor
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10355
positioning technologies in warehouse environments within the scope of the review. Regarding RQ3, key elements of
experiments were derived from the Experimental data in the reviewed articles. An evaluation framework explicitly
tailored for indoor positioning technologies in warehouse environments was established by referencing the house of
quality evaluation system. Figure 3 outlines the review protocol employed in this study, which provides a systematic
methodology for capturing knowledge and insights related to the topic and all relevant variables in a structured manner.
The results of the scoping review based on the proposed review protocol are presented in the subsequent section.
Figure 3. Conceptual framework and review protocol
3. DATA ANALYSIS AND DISCUSSION
In this Section, an analysis of the relevant data from the literature within the review scope was conducted. Statistical
methods, connected papers [22] (a graphical literature search tool that utilizes co-citation and bibliographic coupling
concepts to compile relevant literature lists and graphics), data visualization, and other means were employed to provide
descriptive presentations and analysis of the results. The primary focus of this chapter was on describing the
characteristics of the literature and presenting data relevant to the research questions (RQs).
Figure 4. Publication counts of included articles in the past decade (2014-2023)
3.1 Literature Characteristics
An analysis of literature features can assist researchers in gaining a comprehensive understanding of the status and
development trends in the research field. Initially, the quantity and temporal distribution of the included literature were
analyzed to comprehend the research activity and trends in the application of IPS in warehouse environments. The 38
articles included in this study were published between 2014 and 2023. As observed in Figure 4, the number of publications
has been increasing annually, with significant acceleration noted in recent years. More than half of the articles were
published in the last three years, reaffirming the timeliness and appropriateness of conducting a comprehensive review in
this field. It is worth noting that no articles were identified for the year 2019. This absence results from applying the
2014 2015 2016 2017 2018 2019 2020 2021 2022 2023
0
1
2
3
4
5
6
7
8
9
10
11
Publication Year
LiDRA
Computer Vision
UWB
RFID
VLC
Wi-Fi
Fusion
1
2
4
6
10
11
4
Number of Articl es
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10356
exclusion criteria, where studies not directly relevant to the research objectives, particularly those lacking practical
implications for IPS applications in warehouse environments, were excluded. Next, the keywords and topics of articles
within the research scope were analyzed to identify main research directions, hot topics, and focal points, thus revealing
research themes and trends. Interrelationships among these elements were examined, and a co-occurrence network was
presented in Figure 5.
Figure 5. Keywords co-occurrence network based on the selected 38 articles
In Figure 5, each node expresses a keyword, representing a topic in the field of IPS research in warehouse
environments. The node size indicates the number of articles related to the topic. Nodes are linked by arcs, signifying the
co-occurrence between two topics, and the thickness of these arcs represents the closeness of the subjects. The legend in
the lower right corner of the figure explains that the color of different nodes corresponds to the year of publication of their
related literature, revealing the development trend of popular topics in the field. It is observed from the network that the
primary research objects in this theme are robots (including UAVs), and the main positioning technologies of interest
comprise LiDAR, Computer Vision, RFID, and UWB. A significant increase in warehouse research in recent years is
also noted. In summary, an analysis of the literature has revealed the emerging popularity of the field, but it has not
provided meaningful insights into the application of IPS in warehouse environments. To conduct a comprehensive
investigation, an in-depth analysis and discussion will be centered around the research questions based on the literature
within the scope of this review.
3.2 RQ1: What are the current IPS applications in the warehouse environment?
In the 38 articles included within the scope of this review, applications of IPS in warehouse environments are explicitly
mentioned in 30 articles. In comparison, the remaining 8 articles focus solely on contributions to algorithms without
testing in warehouse application scenarios. Analysis of the survey results reveals that the most frequent area of research
is warehouse inventory applications; 14 out of the 30 articles (46.67%) focus specifically on this area. Since inventory
management is critical to production efficiency and overall economic performance, a significant portion of IPS research
in warehouse environments is allocated to improving these processes. Similarly, warehousing logistics management is
another primary area where IPS is applied, as indicated by 13 articles (43.33%). For the maintenance of efficient
warehouse operations, a well-functioning warehousing logistics management system is deemed essential. Research in this
domain primarily concerns the localization or navigation of mobile devices like AGVs or forklifts within the warehouse
setting. Furthermore, the application of IPS in the digitization of warehouses was discussed in 3 articles (10%). Two of
these articles concentrate on constructing 3D models of warehouses, while the third article focuses on stack measurement
for storing bulk materials. A summary of IPS applications in warehouse environments and corresponding literature is
provided in Table 5. While reviewing applications of IPS in warehouse environments, the contributions and future works
mentioned in the articles under review were also investigated. It was observed that challenges identified in prior research
were targeted for resolution in all scoped articles or enhancements in one or more performance metrics were aimed for.
These articles focus on metrics such as efficiency, safety, accuracy (precision), robustness (stability), and cost. Figure 6
illustrates the distribution of these research articles over the years and their corresponding performance metrics. The
horizontal axis represents the years of publication, while the vertical axis delineates the metrics. Each performance metric
is depicted by a bar graph, where the intensity of the color signifies the relative number of articles published in that
specific period. Darker shades indicate fewer articles, while lighter shades suggest more articles.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10357
Table 5. Applications for IPS in the warehouse
Application areas in the
warehouse Percentage Articles
Inventory
46.67%
[23],[24],[25],[26],[27],[28],[29],[30],[31],[32],[33],[34],[35],[36]
Logistics management 43.33% [37],[38],[39],[40],[41],[42],[43],[44],[45],[46],[47],[48],[49]
Warehouse Digitization
10.00%
[50],[51],[52]
Figure 6. Areas of focus for IPS applications in warehouses (2014-2023)
From the graph, it was observed that some articles address multiple performance metrics simultaneously. A focus on
the positioning accuracy of IPS is seen in a total of 18 articles. A significant increase in articles addressing accuracy-
related issues in the past three years indicates a growing interest in this research area. Conversely, efficiency is the focus
of 8 articles, with a steady growth rate observed year over year. Research on robustness, cost-effectiveness, and safety
has received relatively less attention, although notable trends are still observable. An increase in research on system
robustness since 2017 is evident, whereas hardly any relevant studies existed before that period. Starting in 2020, articles
featuring research on cost-effectiveness and economics appeared. In contrast, a declining trend in publications addressing
safety concerns after 2017 was noticed. Inventorying in a warehouse involves auditing goods, real-time tracking of
inventory, and cross-checking inventory data with financial records to ensure accurate warehouse management. In
traditional warehouse operations, this process relied heavily on manual staff or third-party audits to identify discrepancies
in stock counting, inventory storage, and accounting. With the integration of IPS technologies, inventory management
can be evolved into a highly automated and precise process, enabling real-time updates, reducing human errors, and
enhancing overall efficiency. As illustrated in Figure 7, IPS technologies have been applied across three main warehouse
tasks: inventory management, logistics management, and warehouse digitization. Among these, inventory management
is the most frequently studied application, with the largest share of articles (46.67%) focusing on tasks such as real-time
inventory tracking, cycle counting, and stocktaking. This dominance underscores the critical role of inventory
management in warehouse operations, where precise localization and tracking are essential for operational efficiency.
Logistics management accounts for 43.33% of the studies, primarily focusing on moving and handling goods, such as
internal transportation, warehousing, and order picking. However, many logistics management tasks, particularly those
involving the receipt, storage, and dispatch of goods, are closely tied to inventory management. For example, registering
incoming goods and preparing items for dispatch relies on accurate inventory data and real-time localization capabilities
provided by IPS technologies. In comparison, warehouse digitization comprises only 10% of the articles, focusing on
environmental monitoring and surveillance tasks. While these applications are critical for safety and operational
monitoring, their direct impact on inventory management is less pronounced. From this analysis, it can be concluded that
inventory management represents the largest and most impactful application of IPS technologies and forms the foundation
for many logistics-related tasks. This finding highlights the importance of inventory management when evaluating and
implementing IPS technologies in warehouse environments.
From Figure 7, it was noted that most research related to inventorying tasks utilizes UAVs. This preference is primarily
attributed to the often elevated positions of shelves in warehouse environments, to which UAVs can easily fly, allowing
them to perform real-time inventory checks and stocktaking tasks efficiently. UAVs exhibit distinct advantages in such
scenarios compared to traditional ground-based mobile robots, particularly in high-density storage spaces and multi-level
warehouses. Furthermore, UAVs enable the rapid scanning of inventory and the generation of accurate, up-to-date data
critical for effective warehouse operations. Logistics management tasks within the warehouse, as shown in Figure 7,
primarily involve the internal circulation of goods, including warehousing, material flow, transportation, and express
delivery. In these tasks, mobile devices such as AGVs are predominantly utilized due to their ability to handle payloads
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10358
and navigate efficiently within the warehouse. However, many logistics-related activities, such as receiving, storing, and
dispatching goods, are inherently tied to inventory management processes. These activities rely heavily on accurate
inventory data generated and maintained through IPS technologies, further highlighting the integral role of inventory
management in warehouse operations. Lastly, tasks related to warehouse digitization focus on inspecting and surveillance
personnel, equipment, environment, and goods within the warehouse to detect potential risks and ensure safety. While
these tasks may seem distinct, the real-time data generated by IPS technologies for inventory tracking also contributes to
monitoring warehouse conditions, demonstrating the interconnected nature of these applications. Mobile devices such as
UAVs and AGVs, equipped with advanced IPS technologies, are instrumental across all these tasks, with inventory
management forming the foundation for their broader applications.
Figure 7. Application of IPS in a warehouse environment
In summary, various aspects of warehouse operations, including inventory, logistics management, and warehouse
digitization, are covered by the current applications of IPS in warehouse environments. In inventory and warehouse
digitization tasks, a significant performance has been shown by UAVs, which are widely used. In tasks related to logistics
management, the predominant use of AGVs has been observed. Noteworthy is that the AGVs mentioned in the articles
are equipped with indoor positioning capabilities and the ability to navigate indoors, differentiating them from traditional
AGVs that follow predetermined routes. Therefore, a primary focus in the current applications of indoor positioning
technologies in warehouses is the integration of IPS with mobile robots. Due to their superior space mobility, UAVs are
mainly employed in inventory or inspection tasks. Conversely, logistics-related tasks necessitate mobile devices with
specific payload capacities, resulting in a greater utilization of AGVs for these applications.
3.3 RQ2: What current IPS technologies are utilized explicitly for inventory management in warehouse
environments?
The critical factor of technology choice for determining the suitability of an IPS for a given scenario is addressed in
this section. A comprehensive review of indoor positioning technologies in warehouse environments was first developed.
The technologies identified in the reviewed articles were initially classified, and the methods associated with different
technology categories were subsequently organized and discussed in various research contexts. A deeper understanding
of the relationship between IPS techniques and methods in the warehouse environment was gained by expanding a subset
of the literature using the connected papers method. A comprehensive understanding of the current applications of IPS-
related technologies in warehouse environments is aimed to be provided through these three steps in this chapter.
3.3.1 Classification of technology
In some existing reviews, summaries for IPS have been provided. However, most of these reviews focus on a general
overview of IPS, with little emphasis on its specific application areas. This study initially examined existing
categorizations of IPS technologies from past reviews. Table 6 compares categorization results from three representative
reviews published within the last five years. Subsequently, a comprehensive overview of indoor positioning technologies
applicable to warehouse environments was developed based on the articles included in our scoping review.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10359
Table 6. Classification of IPS technology based on existing review papers (2017-2023)
Reviews
Technology(different)
Technology (similar)
Basiri et al. [53]
Infrared, Magnetometer
Wi-Fi (WLAN)
Tactile Odometer
UWB
Electromagnetic Systems
RFID
Mobile Network
Zigbee
Barometer
Bluetooth (BLE)
Pseudo lite
Visible Light Communication (VLC)
Zafari et al. [54]
None
Acoustic (Ultrasound)
Mendoza-Silva et al. [55]
Tactile Odometer
Computer Vision (Camera)
NFC
IMU
In their Meta-Review on IPS, Mendoza-Silva et al. [55] catalogued various mainstream indoor positioning
technologies, including Light, Computer Vision, Sound, Magnetic Fields, Dead Reckoning, UWB, Wi-Fi, BLE, RFID,
and ZigBee. It should be noted that technologies providing odometer information, such as inertial odometers and wheel
encoders, were not included in their statistical analysis, as they were considered to achieve positioning indirectly. In
contrast, in robot localization, these technologies were found to fundamentally serve the purpose of positioning [56].
Therefore, LiDAR technology was added, and some technologies (such as Bluetooth, Ultrasound, and ZigBee) that did
not appear within the scope of our 38-article review were removed. An overview of the IPS technologies applicable to
warehouse environments is provided in this section and summarized in Table 7. The technologies utilized in the articles
scoped for our review are presented in this table, along with the accuracy reported in some representative studies.
Although existing reviews contain literature that covers a broad spectrum of technologies, only those employed for IPS
are considered in this study. For instance, in the paper by Krug et al. [23], although Computer Vision technology is
mentioned, positioning information was achieved solely through LiDAR technology. Therefore, their positioning
technology has been classified as LiDAR technology. The accuracy metrics presented in Table 7 were extracted from the
articles within our review scope, and only the accuracy from representative research for each technology is displayed.
The use of multi-sensor fusion techniques is also indicated with an asterisk (*) in the table.
As shown in Table 7. In recent years, Wi-Fi positioning technology has been widely applied in warehouse
environments, providing innovative solutions to enhance warehouse automation efficiency. The WiSion system leverages
Wi-Fi signal multipath effects and inertial sensors to estimate a six-degree-of-freedom state in complex indoor warehouse
environments without requiring access point positions, adapting to obstacles and multipath interference [57]. KF-Loc
combines machine learning and Kalman filtering, utilizing millimeter-wave equipment to achieve high-precision
positioning in dynamic warehouse environments, significantly improving smart warehouse management efficiency [58].
The UWB localization method proposed by Zhao et al. [59] achieves centimeter-level accuracy for AGVs without
predefined paths, significantly reducing warehouse automation costs while improving system robustness through data
diagnosis and optimization algorithms. Monica [60] uses UWB technology to achieve high-precision localization for
manual forklifts or personnel. It integrates with laser navigation, greatly enhancing the efficiency of positioning and
managing various equipment and personnel in industrial warehouses. Li et al. [29] combined RFID technology with UAVs
for warehouse inventory management, enabling efficient localization of tagged items on shelves and accurate horizontal
and vertical classification. Alajami et al. [31] proposed the RFID-SOAN navigation system, which uses RFID tags as
digital pheromones to help UAVs autonomously navigate and efficiently perform inventory tasks in mapless warehouses.
Wu et al. [32] introduced the RF-SLAM method, which uses RFID devices to simultaneously localize robots and map
tags, supporting rapid 3D spatial modeling in warehouse environments and enhancing warehouse automation capabilities.
Louro et al. [43] proposed a visible light communication technology applied to warehouse management, enabling
bidirectional communication between infrastructure and autonomous robots and communication among robots. This
system supports robot positioning, transmission of rack information, and interaction on the status of transported items,
enhancing the efficiency of warehouse logistics management. Another application is a VLC-based indoor navigation
system, which uses warehouse LED lighting infrastructure to provide positioning and navigation for AGVs. Through
uplink and downlink communication, automated control of AGVs in warehouse environments is achieved, optimizing the
logistics operations of modern warehouses. The Dual-LiDAR navigation system proposed by Zhang et al. [42]is applied
in warehouses to enable precise autonomous transportation and logistics operations, meeting the demands of intelligent
warehousing with efficient mapping and navigation. The Relative Preintegration (RP) method developed by Kim et al.
[61] enhances the performance of multi-sensor fusion navigation systems, enabling fast and accurate IMU data processing
and improving the adaptability and efficiency of robotic operations in dynamic warehouse environments. Kwon's [27]
system enhanced UAV inventory inspections by ensuring safe navigation in narrow and poorly lit warehouse aisles,
improving operational efficiency. Prakash et al. [49] showcased how leveraging structural features like racks and ceilings
can support precise robot navigation, reducing errors and enhancing automation in large-scale warehouses. Beul et al.
[24] developed an autonomous MAV system that navigates warehouse aisles, identifies stock on shelves, and avoids
obstacles, enabling fully automated inventory inspections guided by a warehouse management system. Gago et al. [51]
designed an aerial robotic system for smart inventory in stockpile warehouses, automating the measurement of bulk
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10360
material volumes, such as fertilizers, with higher accuracy, safety, and efficiency, replacing traditional manual methods
in challenging industrial environments.
Table 7. Summary of indoor positioning techniques in warehouses
Tech
Typical Accuracy
Remarks
References
Wi-Fi
Positioning accuracy with RMSE (root-mean-
square error) is less than 37 cm [58]. Positioning
and orientation errors are 31.77 cm and 2.27◦,
within mild maximum errors and 95%
confidence intervals [57]. Average errors of 89
cm are reported [28].
Despite a more significant error reported
in the article [28], positioning across an
expansive warehouse area is achieved.
[58], [28], [57]
UWB
An average error of approximately 20 cm and a
maximum of around 40 cm are reported [38]. An
average 42.08% reduction in localization error
is noted in three different anchor setups
compared to a baseline approach.
A significant reduction in error is
achieved, but experiments in an actual
warehouse are not conducted [59].
[38],[26], [59],
[60]
RFID
Steady-state error averages no more than 28 cm
[39]. Tracking accuracy ranges from 6 to 10 cm
[29]. Position and orientation RMSE are 15 cm
and 0.2 rad, respectively [63]
*. A mean
accuracy of 13 cm for 3D localization and 0.21
cm for 2D is noted [48]. The mean and standard
deviation of robot l
ocalization via LiDAR
SLAM is 11.9 cm and 5.4 cm [32].
RFID technology is identified as a mature
tag information system extensively used
in logistics and warehousing.
Simultaneous localization of both goods
and mobile robots is achieved using RFID
technology [32].
[39],[27]*,
[29],[63]*,
[31],[48], [32],
[47], [34]
IMU
An average accuracy of 4 cm is reported [42]*.
Both literature [61]* and [42]* discuss the
utilization of IMU for improved
positioning accuracy. However, IMUs are
already integrated into many mobile
robots, such as the Crazyflie nano-
quadcopter mentioned in the literature
[59], which does not address IMU sensor
treatment. Hence, they are not included in
this category.
[61]*, [42]*
VLC
Positioning delay is found to be less than 3ms
[43].
The positioning approach in [43] is
commonly employed to locate target
areas, focusing on latency.
[40], [43]
Computer
Vision
Average errors from ground truth are 3.8 cm for
the proposed method [64]. After more than 60 m
of flight, the final drift
is less than 0.6 m,
equating to around 1% [50].
Average and maximum localization errors are
3.12 cm and 25.68 cm, respectively [27]*. The
average error in X and Y dimensions is less than
5 cm, and the angle is less than 0.1 radian [45]*.
Experimental results and engineering
experiences are comprehensively shared
in the paper [50]. Multi-sensor fusion for
autonomous cargo inventory counting in
real warehouses is accomplished in paper
[27].
[23]*,[25]*,[30],
[35],[37]*,
[64]*,[50],
[61]*,[49]*,
[44],[45]*,[33]
LiDAR
An average positioning error of 11 mm and a
maximum error of 26.18 mm are reported [41].
An average accuracy of 4 cm is noted [42]*. The
mean distance error is 98.2 mm [46].
A comprehensive system, including laser-
based positioning, planned navigation,
obstacle avoidance, and information
acquisition, is presented in the paper [24].
The Omniverse Isaac Sim simulation
environment is employed in the paper
[46]
, enhancing simulation experiment
efficiency compared to traditional
software like Gazebo.
[23]*,[24],
[64]*,[61]*,
[25]*,[52],
[41],[51],
[42]*,[65], [45]*,
[46]
* Note: An asterisk (*) indicates using multi-sensor fusion techniques.
Building upon these advancements, integrating diverse technologies like Wi-Fi, UWB, RFID, VLC, and LiDAR
demonstrates the growing potential for intelligent warehouse systems. These innovations not only improve the efficiency
of logistics operations but also set a foundation for scalable, fully automated warehouse solutions. By addressing
challenges such as navigation in GPS-denied environments, precise inventory management, and real-time communication
between devices, these systems pave the way for smarter, safer, and more adaptable warehouses, meeting the demands of
modern supply chain and logistics industries.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10361
3.3.2 Relationship between technologies, techniques, and algorithms
While classifying the technologies, it was observed that specific techniques or methods corresponded to different
technologies. To delve deeper into the underlying patterns and trends in IPS technology development, the methods
employed to implement each technique were initially categorized based on the articles within the scope of this review.
The relationship between Technologies, Techniques, and Methods is represented in a Sankey diagram, as displayed in
Figure 8. In this diagram, three columns are presented, representing Technologies, Techniques, and Methods,
respectively. Connections between these columns indicate various technical routes. The data visualized in Figure 8 is
derived from the original data summarized in Table 7 of this paper. "Technologies" refers to the different types of
technology utilized in IPS, "Techniques" specifies the unique ways a particular technology is deployed (such as specific
signal attributes or types), and "Methods" predominantly alludes to the algorithms proposed in the research. Additionally,
the label "Sensor Fusion" was added to account for studies that employed sensor fusion techniques for IPS
implementation. Specific algorithm names did not explicitly characterize some optimization methods; these were marked
as "Techniques-based" in their respective studies.
From Figure 8, popular technologies, including LiDAR, Computer Vision, and RFID, were discernible over the past
decade. Preferred techniques, such as Time-of-Flight (TOF), Received Signal Strength Indication (RSSI or RSS), and
Feature Matching, were also identified. In articles within the scope of this review that utilized LiDAR, focus was mainly
placed on the implementation or optimization of IPS. These articles used point cloud data provided by LiDAR for
positioning but omitted details on how LiDAR generated this point cloud data. After this observation, a search was
conducted on LiDAR-related literature. It was found that TOF-based LiDAR is a widely employed technique for distance
measurement in single-point depth sensing and 3D mapping [68]. Therefore, most research on LiDAR technology in this
context utilizes TOF-based distance sensing, with variations in the algorithms employed for data processing. This Sankey
diagram systematically illustrates the relationships between Technologies, Techniques, and Methods, clearly visualising
how different technical pathways are applied in IPS research. The diagram is divided into three columns: the left column,
Technologies, includes various types of technologies such as IMU, VLC, Wi-Fi, UWB, RFID, Computer Vision, and
LIDAR. The middle column, Techniques, represents the specific application methods of these technologies, such as
Signal-to-Noise Ratio (SNR), Angle of Arrival (AOA), Received Signal Strength (RSS), Time Difference of Arrival
(TDOA), Visual Odometry (VO), and feature extraction. The right column, Methods, summarizes the algorithms and
solutions proposed in related studies, including machine learning approaches, optimization methods, and fusion strategies.
Figure 8. Relationship between technologies, techniques, and methods of IPS in warehouse environments
The connections between the columns illustrate the pathways from Technologies to Techniques and finally to
Methods, showing how various technologies are applied to specific algorithms through techniques. For example, Wi-Fi
technology is primarily associated with RSS and SNR techniques, which connect to multiple positioning algorithms.
UWB technology is often linked with TDOA and trilateration methods, contributing to path planning and fusion strategies.
Sensor Fusion is specifically highlighted, reflecting its cross-technology and multi-method applications. While not
explicitly labeled with algorithm names, some optimisation approaches are summarized as “Techniques-based” pathways.
It is important to note that the Methods listed on the right side of the diagram are derived from the original data
summarized in Table 7 of this paper. By visualizing the data from Table 7, this Sankey diagram provides a comprehensive
overview of the interconnections between technologies, techniques, and algorithms in existing research. It delivers a clear
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10362
analytical framework for understanding these relationships and serves as a valuable reference for future research in
selecting technologies and planning technical pathways.
3.3.3 IPS technology development in warehouse environment
In the previous section, the organization of technologies, techniques, and methods relevant to IPS was carried out
within the scope of the review, and their relationships were identified. Despite the review being limited to 38 articles,
various technologies were observed. Consequently, challenges were faced in providing a comprehensive and clear
background for these technologies, techniques, and methods based solely on the articles within the review. A methodology
involving the search for secondary and tertiary literature referenced in the primary articles was employed to address this
issue. This approach enhanced the technology framework and gave readers a more comprehensive and detailed field
review. The 38 primary articles included in the review and the classification results from Table 8 were input as original
data into the connected papers system [22], and secondary and tertiary literature related to each technology was
systematically identified. Literature that cited each technology extensively was meticulously examined, and an expanded
review library of relevant literature was subsequently assembled. The literature thus discovered is presented in Table 8.
Table 8. Ultimate tensile strength values and elongation to fracture
Tech
Tracking Literature
Wi-Fi [66], [67], [68], [69], [70], [71], [72], [73], [74], [75], [58], [76]
UWB
[77], [78], [79], [80], [81], [82], [83], [84], [85], [86], [87]
RFID
[88], [89], [90], [91], [92], [93], [94], [95],
VLC
[96], [97], [40], [43]
IMU
[98], [99], [100], [101], [61]
Computer Vision
[102], [103], [104], [105], [106], [107], [108], [109], [110], [111], [112], [113], [114]
LiDAR
[115], [116], [117], [118], [119], [120], [121]
Seven different types of IPS-related technologies were included in the review. These technologies were classified into
two types based on the measurement medium: wireless signals, such as Wi-Fi, UWB, and RFID, and other physical
signals, like Computer Vision, LiDAR, IMU, and VLC. Algorithms for indoor positioning technologies based on radio
signals were categorized into AOA [70], Time of Arrival (TOA) [67], [68], TDOA [69], and Received Signal Strength
Indication (RSSI) [71]. These algorithms include geometric localization methods like triangulation and trilateration, as
well as adjacency information and fingerprint localization methods such as adjacency-based positioning, multiliterate,
and fingerprint recognition [85]. The algorithms associated with the other four types of physical signals (Computer Vision,
LiDAR, IMU, VLC) used for IPS implementation were also summarized.
Despite the extensive research on IPS technologies in warehouse environments, several critical gaps remain. First,
while many studies emphasize technology adoption, there is limited focus on how these technologies address the specific
challenges of dynamic inventory management in large-scale warehouses. Second, the trade-offs between cost, precision,
and scalability are often overlooked, leading to a lack of practical guidance for technology selection. Third, few studies
explore hybrid solutions combining multiple IPS technologies to balance their strengths and limitations. These gaps
underscore the need for tailored evaluation frameworks and innovative approaches to optimize IPS deployment for
inventory management.
Table 9. Overview of common wireless localization methods and principles
Methods
Described
References
Geometric
Measurement
TOA
Determination of position by measuring the propagation time of a
signal from transmitter to receiver.
[67]
TDOA
Determination of position by measuring the time difference between
signal arrivals at multiple receivers with known positions.
[69]
AOA
Determines position by measuring the AOA of the signal using an
antenna array or directional antenna on the receiver.
[70]
Fingerprinting
RSSI
(RSS)
The fingerprint localization method establishes a correspondence
between the geographic location of each point in the indoor space and
the signal. It achieves positioning through feature matching.
[70], [71]
3.3.3.1 Wi-Fi-based IPS technology
The principle of active positioning in Wi-Fi-based positioning is established by placing a certain number of Access
Points (APs) in the indoor environment. When the mobile receiving end enters the positioning area, a search for the APs
to transmit wireless signals is initiated, and location is determined through the received signal values, as shown in
Figure 9. Currently, mainstream Wi-Fi positioning technologies fall into two categories: active Wi-Fi localization
technologies that utilize geometric measurements and passive positioning technologies based on Wi-Fi fingerprint
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10363
information [66]. The categorization of Wi-Fi-based IPS technologies is presented in Table 9. In geometric measurement-
based Wi-Fi positioning technology, the distance from the receiving device to each wireless AP is calculated, and the
target position is determined through distance intersection. However, the accuracy of distance measurement using Wi-Fi
signals is compromised, as Wi-Fi signals are not explicitly designed for positioning. The performance of the Wi-Fi
positioning method based on geometric measurement is negatively affected in indoor environments due to signal
multipath, reflection, and refraction. Additionally, Wi-Fi positioning technology is susceptible to variations in signal
strength. Active positioning schemes that rely on single wireless technologies, such as Wi-Fi, UWB, and RFID, include
TOA [67], TDOA [69], and AOA [70], as shown in Figure 10.
Figure 9. Schematic diagram of Wi-Fi-based IPS principle
(a)
(b)
(c)
Figure 10. Active wireless location method: (a) TOA, (b) TDOA, and (c) AOA
The measurement object of both TOA and TDOA methods is the time signals are disseminated. By multiplying the
speed of the signal by the TOA data, the relative distance between the signal source and the measurement point can be
calculated. Therefore, the crux of the method lies in the accurate acquisition of signal propagation time. A mechanism to
improve TOA or AOA localization performance by transmitting multiple predefined messages was proposed by Yang et
al. [68], allowing for reduced network bandwidth and antenna requirements while maintaining high-accuracy
performance. In a study by Cheng et al. [72], the Taylor algorithm was developed, achieving 1-decimeter positioning
accuracy in indoor line-of-sight (LOS) environments with dynamic and static data. The AOA-based ranging positioning
method relies on the angle between the target and at least two known positions to pinpoint the target's location. A robust
phased array-based positioning system, which adopts a sparse reconstruction algorithm to improve AOA algorithm
accuracy significantly, was proposed by Gong et al. [73]. In 2020, Vashist et al. introduced an indoor warehouse location
system using a 60GHz wireless router and SNR as a feature of consumer-grade wireless APs in a machine learning-based
location algorithm. The system achieved Remarkable centimetre-level accuracy with an RMSE of 0.84m and an MAE of
0.37m, meeting the accuracy requirements for warehouses [58].
In the realm of passive Wi-Fi positioning technology based on fingerprint information, the following concept applies:
Wi-Fi fingerprint localization is predicated on the correspondence between the geographic location of each point in indoor
space and the signal, achieving positioning through feature matching, as depicted in Figure 11. Indoor environments are
partitioned into several blocks, each possessing a unique "fingerprint," which encapsulates the characteristic information
of the location and other features that constitute part of the "fingerprint library" [68]. Two signals are primarily relied
upon in fingerprint localisation methods: RSSI and CSI. RSSI represents a quantized measurement of the physical signal
strength received, known as RSS, and is utilized in various wireless applications. CSI provides a more comprehensive
assessment of channel conditions between the transmitter and receiver. The establishment of an accurate channel
attenuation model is identified as critical for RSSI-based ranging and positioning algorithms. A wireless map with fine-
grained CSI was established by Shi et al. [74] to improve target location estimation accuracy. The SpotFi system, proposed
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10364
by Hoang et al. [75], utilized a convolutional neural network (CNN) to train signal features, thereby enhancing the
accuracy of RSSI measurement. A deep learning model equipped with an attention module was employed by Brunello et
al. [76] for the first time, improving fingerprint-based IPS in both theoretical aspects and localization accuracy, achieving
a range of 0.8-1.0 m in accuracy.
Figure 21. Schematic representation of the fingerprinting localization method for Wi-Fi-based indoor positioning
In summary, geometry-based active Wi-Fi positioning techniques are constrained by their low accuracy in indoor
environments affected by signal multipath, reflection, and refraction, as well as their susceptibility to fluctuations in signal
strength. Based on the literature surveyed, the challenge lies in obtaining accurate signal propagation times through
improved algorithms or device performance. Passive Wi-Fi positioning techniques based on fingerprint information face
limitations, such as the significant effort required to build an offline fingerprint database and the difficulty adapting to
environments undergoing substantial changes. Establishing an accurate channel fading model is a challenge for RSSI-
based ranging and positioning algorithms.
3.3.3.2 UWB-based IPS technology
The US Department of Defense first proposed UWB technology in the 1960s, and it was primarily utilized for military
applications at that time [77]. It wasn't until 1998 that the Federal Communications Commission (FCC) authorised civil
use of the technology. UWB technology utilizes an extensive frequency band, ranging from 3.1 to 10.6 gigahertz, by
transmitting very short pulses, thus providing significant bandwidth advantages and short pulse periods. As a result, UWB
can offer greater capacity and higher data rates. In addition, it performs well in low signal-to-noise ratio communication
channel conditions and is immune to multipath propagation conditions. This makes UWB communications suitable for
indoor positioning applications [78], especially in non-line-of-sight (NLOS) conditions. Since the transmission is in short
pulses, UWB signals are transmitted with a low average power spectral density, placing them on the noise floor (typically
-40 dBm/MHz), resulting in reduced transmit power consumption, improved power efficiency, and resistance to
interference and interception, as shown in Figure 12 [122].
Figure 12. Comparison of UWB spectral properties with various positioning techniques [122]
UWB technology shares similarities with Wi-Fi regarding technical routes, as both utilize wireless signals for
positioning. However, their characteristics and application scenarios differ. Like UWB, Wi-Fi can achieve positioning
through methods such as RSSI, TOA, or AOA, as shown in Figure 10 and Figure 11. Nevertheless, UWB employs ultra-
wideband frequencies and short-pulse transmission, providing higher precision and stronger resistance to multipath effects
than Wi-Fi's narrowband communication, especially in non-line-of-sight (NLOS) environments. Meanwhile, Wi-Fi, with
its widespread infrastructure deployment and lower costs, is better suited for large-scale indoor coverage scenarios. Thus,
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10365
while UWB and Wi-Fi share overlapping principles, the former focuses more on high-precision and low-latency
positioning requirements, whereas the latter emphasizes accessibility and cost efficiency. The two technologies exhibit a
complementary relationship in indoor positioning applications.
In UWB-based geometric measurements, the primary focus of researchers has been on error reduction in NLOS
environments, where obstacles can obstruct the direct path between the transmitter and receiver, leading to signal
reflection, diffraction, or scattering. A machine learning-based algorithm for classifying signal propagation in LOS and
NLOS situations was developed by Marano et al. [79], effectively reducing the resulting errors. The variation law of
residuals between LOS and NLOS indoor environments was studied by Zhang et al. [80] using the Kalman Filtering
algorithm (KF). NLOS errors were identified and mitigated by setting an appropriate threshold and comparing real-time
residuals with ranging. Yu et al. [81] corrected signal arrival time estimation when UWB signals propagated in complex
indoor NLOS environments. A method for NLOS error elimination based on TOA was proposed by Go et al. [82], where
the distance measurement value of the signal propagation between different base stations and tags performs the distance
compensation. An algorithm that combines the least squares method and the Kalman Filter algorithm was introduced by
Wang et al. [83], demonstrating a positioning accuracy of 9.3 cm. Research into UWB-based fingerprint positioning
methods has included techniques to reduce positioning errors from NLOS and multipath effects. Channel measurement
[85], front-end energy sampling on the receiving end [86], and measurement error noise are commonly utilized techniques.
A new UWB positioning system based on an unmanned aerial vehicle (UAV) using integrated Radio Frequency (RF)
hardware and antennas was proposed by Tiemann et al. (2015) [87]. The challenges of achieving accurate UAV
localization in large warehouses were discussed by Macoir et al. [84], especially considering the high costs associated
with deploying complex wiring and power infrastructures required by large-scale UWB systems.
High measurement accuracy is one of the main advantages of UWB positioning technology, with location errors
reported to be less than 10cm in some instances [83]. Good multipath mitigation is also offered by UWB, making it
suitable for environments with high-density tags and high mobility. However, limitations of UWB positioning technology
include the time-consuming and expensive initial setup, which requires precise calibration and placement of transceivers,
and the need for an unobstructed path between the transmitter and receiver. These limitations can compromise its
effectiveness in specific indoor environment settings. UWB is widely used in warehouse environments for inventory
tracking, AGV navigation, and real-time personnel monitoring. For instance, UWB-based positioning systems have been
implemented to enable high-precision tracking of goods on high shelves, ensuring inventory accuracy while minimizing
manual intervention. Moreover, UWB's ability to provide robust performance in high-mobility scenarios has been
leveraged for AGVs, where its low latency and high accuracy are critical for collision avoidance and optimal path
planning.
UWB technology excels in indoor positioning with high accuracy, strong NLOS performance, and resistance to
multipath effects, making it ideal for inventory tracking and AGV navigation tasks. However, it faces challenges such as
high deployment costs, complex calibration, and scalability issues in large-scale environments. To overcome the inherent
limitations of UWB, researchers have explored its integration with other positioning technologies. For example, UWB-
LiDAR fusion systems have enhanced 3D mapping and navigation accuracy, particularly in cluttered indoor environments
[60]. Although these hybrid systems improve performance, they also increase complexity and costs. Future advancements
in UWB technology focus on simplifying deployment and reducing costs by utilizing software-defined systems equipped
with real-time calibration and adaptive algorithms. Additionally, integrating AI is anticipated to enhance NLOS error
mitigation and boost signal reliability, further broadening UWB’s robotics, smart warehouses, and healthcare
applications.
3.3.3.3 RFID-based IPS Technology
Derived from the rapid development of radar communication technology in the 1950s, RFID is a non-contact method
for transmitting information. Utilizing RF signals through spatial coupling, this technology serves the purpose of
automatic identification [88]. Over the years, substantial advancements have been made in the technical theory of RFID.
Initially applied in the field of indoor positioning since the start of this century, RFID has given rise to classic positioning
systems like the SpotON [89], LANDMARC system [90], and VIRE system [91]. As shown in Figure 13, RFID tags are
deployed on the ground in a specified pattern for RFID-based positioning. The ID information of an RFID tag is tied to
its coordinate position on the ground, so RFID-based e-maps can be defined based on the tag's ID and position [123]. The
LANDMARC system operates by deploying a network of fixed reference points, such as access points or RFID readers,
throughout indoor areas. These reference points are anchors for measuring the wireless signals emitted by mobile devices
or tags carried by individuals or objects. Several modifications and enhancements to the LANDMARC system have been
proposed to address the complexity of indoor environments. Gu et al. (2020) introduced a novel indoor positioning
algorithm based on LANDMARC to balance cost and precision. The system replaced physical tags with "reference tags,"
reducing electromagnetic interference and system costs. Furthermore, Hu et al. [90] proposed an optimized LANDMARC
positioning algorithm to mitigate significant differences in RSS between tags situated close to the reader [92].
However, conventional methods like LANDMARC and K Nearest Neighbors (KNN) often suffer from limited
accuracy due to signal reflection, diffraction, and non-occlusion factors. Zhou et al. [93] presented an improved KNN
method that corrected target coordinates using a passive RFID system. Subedi et al. [94] achieved centimeter-level
localization accuracy in complex environments using only RSSI measurements from multiple passive tags. Li et al. [95]
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10366
introduced the WIMEC-LANDMARC algorithm that incorporates average error correction, improving accuracy.
Additional advancements include work by Teo et al. [39], who demonstrated effective autonomous mobile robot
navigation using RFID signal strength sensing. Tao et al. [123] proposed a Monte Carlo and dual-antenna joint corrective
fit-based scheme, which showed significantly higher localization accuracy than particle filter-based algorithms. A recent
trend involves the incorporation of intelligent algorithms into RFID-based IPS. Notably, a neural network-based
optimization method was proposed, significantly improving system localization accuracy [95].
Figure 13. RFID-based Indoor positioning system
3.3.3.4 VLC-based IPS technology
Visible Light Positioning has emerged as a dual-functional technology, offering illumination and communication
features. This innovative approach is gaining significant traction in the field of IPS [96]. Specifically, VLP, a subset of
VLC, is identified as an up-and-coming solution for indoor positioning, as shown in Figure 14 [97]. The system is a high-
precision indoor positioning solution based on a single LED lamp. It consists of an original VLP lamp paired with a small
luminescent beacon mounted on its edge, which emits encoded visible light signals. Using a CMOS sensor with a rolling
shutter mechanism, the system captures bright and dark stripes formed by the light signal. Image processing algorithms
extract the pixel coordinates of the lamp and beacon, which are combined with the beacon's physical coordinates to
calculate the precise position of the device using trigonometric functions. A beacon-searching algorithm further
accelerates the localization process. The low-complexity design requires processing only a single set of bright and dark
stripes without binarization or complex projections, ensuring high efficiency and low hardware requirements. The system
achieves centimeter-level accuracy (average error of 2.26 cm) and millisecond-level response times (average positioning
time of 6.3 ms), making it suitable for indoor navigation applications on low-cost embedded platforms.
Figure 14. VLC-based IPS system architecture [97]
In recent years, Chen et al. [124] proposed a method that employed fingerprinting and an Extreme Learning Machine
(ELM) to achieve high localization accuracy, robust interference immunity, and excellent real-time performance. This
method reported an average 3D positioning error of just 2.11 cm. Following this, Li et al. [96] presented an unbalanced
single LED VLP algorithm and a fast beacon search method. For indoor settings with a height of 3 meters, their approach
yielded a positioning accuracy of 2.26 cm. Practical applications of VLC in warehouse environments were researched by
Louro et al. [40] [43]. A white LED lighting system was installed on the warehouse ceiling to enable bidirectional
communication between the infrastructure and vehicles. This system comprises two core elements: VLC transmitters,
LED lights, and AGVs with VLC receivers. Data is transmitted by white RGB LED emitters in the LED lights and
collected by VLC receivers on the AGVs. An ON-OFF keying method is employed for data modulation [40]. Three
primary color white LEDs and photodetectors are utilized for the transmitters and receivers. Each LED light provides
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10367
positional information to the vehicles through the adequate modulation of RGB emitters. Reliable data transmission is
ensured by coding and synchronization techniques, and error detection and correction are enhanced through parity bits
[43].
Table 10 highlights three key VLC-based indoor positioning methods, each leveraging unique techniques to achieve
high precision and real-time performance. The Single Lamp and Beacon System stands out for its simplicity and low
hardware requirements, making it suitable for low-cost embedded platforms, with an accuracy of 2.26 cm and a response
time of 6.3 Ms. The Fingerprinting + ELM Method achieves the highest reported 3D positioning accuracy of 2.11 cm by
using machine learning to enhance robustness and real-time performance, making it ideal for static environments requiring
high precision. Similarly, the Asymmetric Single Lamp VLP Algorithm focuses on fast localization with simple hardware
design, balancing cost and efficiency. The Bidirectional Communication and Navigation System integrates positioning
with communication, facilitating warehouse AGV navigation by leveraging RGB LED lights and robust error correction
techniques. The key advantages of VLC-based IPS include its compatibility with existing lighting infrastructure in indoor
spaces, making VLP highly convenient. Additionally, in environments where wireless signals are susceptible to
interference, such as hospitals with MRI equipment or factories with large electromagnetic devices, VLC-based IPS is
highly applicable. VLC technology also demonstrates less susceptibility to multipath effects and interference from other
wireless systems. Nevertheless, challenges are faced by VLC-based IPS, particularly in maintaining positioning accuracy
in environments where light reflections, diffusion, or dynamic obstacles are present. Recent advancements in VLC
technology have focused on improving signal processing using machine learning algorithms. For instance, convolutional
neural networks (CNNs) have been applied to enhance the robustness of signal decoding under conditions of high ambient
light interference. Additionally, hybrid systems combining VLC and Wi-Fi have been proposed, leveraging the high-
speed communication capabilities of VLC with the broad coverage of Wi-Fi to create a complementary system for large-
scale indoor positioning.
Table 10. Overview of visible light positioning techniques and their performance
Method
Description
Performance
Features
Single Lamp and
Beacon System [96]
Utilizes a single LED lamp paired with
a small beacon. A CMOS sensor
cap
tures bright and dark stripes, and
trigonometric functions calculate the
device's position.
Accuracy: 2.26 cm
Response time: 6.3
Ms
Simple design, low hardware
requirements, suitable for low-
cost embedded platforms.
Fingerprinting + ELM
[124]
Matches light signal features with a pre-
built fingerprint databa
se. Combines
Extreme Learning Machine (ELM) to
enhance accuracy and interference
immunity.
3D Positioning
Error: 2.11 cm
It relies on environment
calibration, is
suitable for
fixed scenarios, and enhances
stability via machine learning.
Bidirectional
Communication and
Navigation System
[43]
It uses RGB LED lights to modulate and
transmit position information. AGVs
with VLC receivers enable bidirectional
communication and navigation.
Accuracy: Not
specified
Combines communication and
positioning, high reliability,
suitable for ware
house AGV
navigation.
3.3.3.5 IMU-based IPS technology
Inertial Measurement Unit (IMU) positioning is a technology that achieves three-dimensional spatial position and
orientation estimation through sensor fusion and is widely used in drones, autonomous driving, robotics, and indoor
navigation. It relies on accelerometers, gyroscopes, and magnetometers, integrating sensor data using Direction Cosine
Matrix (DCM) or quaternion methods to calculate attitude angles in real-time while correcting drift errors [98].
Accelerometers provide linear acceleration data for displacement calculation, gyroscopes measure angular velocity to
estimate rotation, and magnetometers reference the Earth's magnetic field to determine heading. To address cumulative
errors caused by traditional double integration, jerk integration is employed for displacement calculation, and Extended
Kalman Filter (EKF) further refines multi-sensor state estimation. Barometric pressure sensors or laser rangefinders assist
in altitude measurement, enabling comprehensive 3D localization [99]. Despite its low cost and operational convenience,
indoor positioning technology based on inertial navigation is subject to inevitable accumulated errors and requires per
iodic calibration through external information. Consequently, research focused on attitude update algorithms for this
system has garnered considerable attention. The Pedestrian Dead Reckoning (PDR) algorithm, a method that uses inertial
sensors to calculate the distance and direction of a target's movement, was first proposed by Wu et al. [100]. This method
enables the calculation of the target's relative position. Figure 15 illustrates the general framework of an IMU-based
Pedestrian Dead Reckoning (PDR) system, providing a comprehensive depiction of the process from sensor data
collection to trajectory output. The system is centered on accelerometers, gyroscopes, and magnetometers, which measure
linear acceleration, angular velocity, and geomagnetic direction, respectively, with barometers optionally used for altitude
variation measurement. In the preprocessing stage, features are extracted through gait detection and motion classification.
Gait detection leverages methods such as Zero-Velocity Update (ZUPT) and Zero-Angular Rate Update (ZARU) to
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10368
identify gait phases. At the same time, motion classification employs machine learning algorithms to distinguish motion
types (e.g., walking, running), ensuring high-quality input for trajectory calculation.
Figure 15. IMU-based pedestrian dead reckoning system framework [100]
The core algorithms include Inertial Navigation Systems (INS) and Step-and-Heading Systems (SHS). INS estimate
positions through integration while combining ZUPT and EKF to mitigate error accumulation. SHS calculates trajectories
by incrementally adding step lengths and headings. Map matching and magnetic field correction also constrain trajectory
errors, enhancing accuracy. Ultimately, the system outputs high-precision three-dimensional positions, trajectories, and
orientation information. Through modular design, data fusion, and environmental constraints, this framework effectively
addresses error issues in inertial navigation, providing a reliable solution for indoor positioning. Recent advancements in
IMU-related algorithms have focused on improving positioning accuracy in dynamic and cluttered warehouse
environments. For instance, Tong et al. [101] proposed an Enhanced PDR algorithm optimized for AGV navigation in
high-density storage areas, achieving a 15% reduction in cumulative error compared to traditional methods. To mitigate
the cumulative error inherent in IMU-based positioning, hybrid systems integrating IMU with Simultaneous Localization
and Mapping (SLAM), LiDAR, or vision-based technologies have gained significant traction. For example, in a study by
Qin et al. [113], the VINS-Mono system combined IMU with monocular vision to achieve precise localization in dynamic
warehouse environments, demonstrating long-term operational stability with a maximum error of 8 cm under the EuRoC
dataset. Similarly, Kim et al. [61] proposed a method for calibrating IMU and LiDAR pairings, enabling robust navigation
in cluttered warehouse spaces. These hybrid systems leverage the complementary strengths of IMU and other sensors,
significantly improving positioning accuracy and reliability in complex indoor environments.
In conclusion, IMU-based indoor positioning is widely adopted in navigation and positioning equipment due to its
cost-effectiveness, compact design, and ability to operate independently of external signals. This autonomy makes IMU
systems particularly suitable for complex warehouse environments characterized by narrow aisles, high-density storage,
and dynamic obstacles, where reliable and precise localization is critical. The development of Micro-Electro-Mechanical
Systems (MEMS) has further enhanced the precision and reliability of IMU-based systems in indoor settings. However,
challenges such as cumulative errors and sensitivity to external disturbances necessitate periodic calibration and
algorithmic improvements. To address these issues, researchers are focusing on adaptive filtering techniques, enhanced
sensor fusion methods, and the integration of IMU with technologies such as SLAM and LiDAR, which have
demonstrated significant potential in improving positioning accuracy and robustness. Future advancements in real-time
data fusion and error correction algorithms are expected to expand the applicability of IMU-based systems, making them
a vital component of intelligent warehouse operations.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10369
3.3.3.6 Computer vision-based IPS technology
Computer vision technology was developed with advancements in sensor devices, computer computing power, and
image processing technology. Since the 1980s, machine vision technology has been extensively researched. Techniques
for vision-based positioning, including monocular vision, binocular vision, and RGB-D, are included. The process flow
for obtaining environmental image information through the camera lens in computer vision positioning technology is
illustrated in Figure 16. Firstly, images are captured by a calibrated camera [125], then image processing and analysis are
performed, and finally, the required information about the external environment is derived. Computer vision-based IPS
technology can be classified into known and unknown environments according to the prior information of the environment
possessed by the receiving device [102].
Figure 16. Process flow of computer vision positioning technology
In known environments, the location of equipment is established based on image data captured by a camera using
visual positioning technology. Previously mapped environments with features such as artificial beacons or markers are
referred to as known environments. During the localization process, the captured image is matched with images or markers
stored in a database and location information is obtained from the most similar image. The first introduction of computer
vision for indoor positioning was made by a mobile robot designed by Kriegman et al. [103]. Linear geometric features
were extracted from images captured by a monocular camera on the robot, and the extended Kalman filter was used to
reduce uncertainty and determine the camera position. Since then, various approaches for optimizing feature extraction
methods and improving localization performance have been proposed. Simultaneous Localization and Mapping, a widely
used technique in robotics and indoor navigation, has further advanced the integration of computer vision into indoor
positioning systems. SLAM enables a device to construct a map of an unknown environment while simultaneously
localizing itself within that map. The SeqSLAM method, proposed by Milford et al. [104], uses image differences to
measure similarity between images and improves localization accuracy through graph matching. Vehicle speed and
distance were estimated by an algorithm presented by Ho et al. [105] utilizing a monocular camera to measure optical
flow and control inputs. The robustness of image recognition was enhanced by a place recognition method based on LDB
(Local Difference Binary) features proposed by Arroyo et al. [106]. A neural network called NetVLAD for extracting
image features was proposed by Arandjelovic et al. [107], using a large amount of image data to learn representations for
VLAD (vector of locally aggregated descriptors) features [108].
In unknown environments where existing beacon methods are not applicable, the environment is reconstructed through
real-time and online video. The position of the image sensor is calculated in real-time using Visual Simultaneous
Localization and Mapping technology [102]. Feature point associations between two images are established by extracting
and matching image feature points. Peripolar geometry is used to solve camera motion, and triangulation is used to
calculate the 3D information of features [109]. Davison developed a SLAM technology positioning system based on
monocular vision in 2003, combining the SLAM algorithm with visual positioning technology [126]. ORB-SLAM, a
robust positioning system for locating vigorously moving targets, was proposed by Mur-Artal et al. [110]. Support for
different vision devices and fully automatic initialization were added in the ORB-SLAM2 [111] and ORB-SLAM3 [112]
systems, developed in 2017 and 2021, respectively. The components of ORB-SLAM1, ORB-SLAM2, and ORB-SLAM3
are depicted with different background colors to indicate their version-specific functionalities in Figure 17. ORB-SLAM1
centers on TRACKING, LOCAL MAPPING, and LOOP CLOSING modules for basic navigation and mapping.
Additional features in ORB-SLAM2 include support for more camera types and enhanced accuracy [111]. ORB-SLAM3
introduces IMU integration and the Atlas module for advanced map construction and localization, increasing accuracy
across varied scenes.
Hardware constraints and the ever-expanding range of positioning targets impose significant limitations on the
localization accuracy of pure visual SLAM. Consequently, the development of multi-sensor positioning systems that
incorporate vision technology, inertial navigation, LiDAR, and wireless communication technology has been deemed
necessary. Such a system is the VINS-Mono, developed by Qin et al. [113], in which localization is achieved by
integrating vision and inertial navigation IMU devices. High stability, contributing to long-term operational accuracy and
robustness, is exhibited by the VINS-Mono and VIN-Mobile systems. Under the EuRoC data set, a maximum error of 8
cm in the localization accuracy of the VINS-Mono system is recorded. A system that combines computer vision and Lidar
was proposed by Mohta et al. [50], enabling rapid and reliable autonomous navigation even with limited prior
environmental knowledge. This system achieves a maximum speed of 7 m/s and a final position drift of less than 2 m.
The lowest-cost navigation platform for unknown cluttered environments to date has been implemented by Campos-
Macías et al. [114] using the Intel Ready to Fly drone kit. Overall, computer vision-based indoor navigation technology
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10370
promises accurate and real-time positioning in indoor environments. One of its main advantages is that no additional
infrastructure or hardware installation is required, rendering it cost-effective and easy to deploy. Suitability for low-light
conditions is another advantage. However, drawbacks exist, such as the requirement for high computational power, which
may limit real-time performance. Additionally, accuracy may be influenced by the quality and quantity of visual features
in the environment.
Figure 17. The main system components of ORB-SLAM1-3 and their enhancements
3.3.3.7 LiDAR-based IPS technology
LiDAR (Light Detection and Ranging) is identified as a sensing technology in which lasers are employed to measure
distances and construct precise 3D models of an area. Its capacity to operate in GPS-denied and weak signal indoor
settings has contributed to its broad adoption for indoor positioning due to its accuracy and reliability in recent years
[115]. In LiDAR indoor positioning technology, laser beams are emitted in all directions by a LiDAR sensor. These beams
are bounced back upon encountering objects, and the Time-of-Flight between emission and return is analyzed to calculate
the object's distance and position. This process is repeated multiple times, culminating in a large point cloud data set that
represents the 3D environment. Technology has facilitated integration with the specific requirements of various fields,
leading to its extensive application in diverse domains [116].
SLAM utilizing LiDAR technology has been recognized as a significant research direction in mobile robotics [111].
As illustrated in Figure 18, the basic framework of a LiDAR-based SLAM system consists of several key components
that work together to achieve accurate localization and mapping. First, “data sensing” is handled by the LiDAR sensor,
which emits laser beams to capture point cloud data representing the surrounding environment. This raw data is processed
in the next stage, where “data processing and estimation” are managed by the odometer. The odometer estimates the
sensor's position and orientation by aligning successive scans using Iterative Closest Point (ICP) algorithms.
Subsequently, “global map construction” optimizes the alignment of multiple scans across a larger area, correcting
cumulative errors and ensuring consistency in the 3D map. Finally, “loopback detection” identifies previously visited
locations, aligning them with current scans to eliminate drift and enhance overall map accuracy, particularly during
extended mapping sessions. Among its advantages is the technology's functionality in GPS-compromised environments,
such as warehouse environments. LiDAR-based IPS technologies are widely utilized for inventory management,
automated guided vehicle navigation, and high-density shelving structural inspection. Generating precise 3D models
enables LiDAR-equipped AGVs to navigate complex warehouse layouts accurately, avoiding obstacles and optimizing
path planning in real-time. LiDAR sensors mounted on UAVs have also been employed for inventory inspections in
elevated storage areas, providing centimetre-level accuracy in locating and auditing goods. These applications highlight
LiDAR's unparalleled ability to address the challenges of high-density, dynamic warehouse environments.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10371
Figure 18. Basic framework for SLAM
Several challenges exist in employing LiDAR for indoor positioning, including data processing and noise filtration.
Various researchers have tried to enhance the accuracy of LiDAR for indoor positioning. A method combining motion
models with LiDAR measurement data was proposed by Sánchez et al., which uses infrastructure elements as positioning
references. Wang et al. [117] presented an efficient algorithm that employs LiDAR as the sole environmental detection
sensor in IPS research, thus reducing computational effort while preserving localization robustness [118]. Recent
advancements in LiDAR-based IPS systems have focused on improving point cloud data processing and reducing
computational overhead. For instance, Shi et al. [119] proposed a lightweight SLAM algorithm that achieves real-time
performance in high-density storage environments, reducing latency by 30% compared to traditional SLAM methods.
Hardware innovations, such as the development of low-cost solid-state LiDAR sensors, are also expanding the
accessibility of LiDAR technology in cost-sensitive warehouse applications. Two commercially available solid-state
LiDAR SLAM frameworks with mature technology currently exist: the FAST LOAM framework [120], and the LOAM
Livox framework [121]. A multi-sensor fusion framework was put forth by Kwon et al. [25] to facilitate practical
autonomous UAV navigation in GPS-deprived, poorly lit warehouses, albeit at a relatively high cost.
Future research on LiDAR-based IPS systems is expected to explore further integrating artificial intelligence (AI)
techniques to enhance system adaptability and efficiency. Deep learning algorithms can process point cloud data, enabling
more accurate object recognition and anomaly detection in warehouse environments. Additionally, the combination of
LiDAR and predictive analytics may facilitate proactive maintenance and inventory forecasting, paving the way for fully
autonomous and intelligent warehouse operations. In summary, the analysis of IPS technologies highlights their
transformative impact on inventory management within warehouse environments. With its scalability and cost-
effectiveness, RFID is widely used for ground-level inventory tracking and bulk item identification. UWB stands out for
its high precision and adaptability in dynamic inventory tasks, ensuring real-time updates even in complex and densely
packed storage areas. LiDAR offers unparalleled 3D mapping capabilities, particularly suited for high-density and
elevated inventory spaces. At the same time, Wi-Fi provides an accessible and cost-sensitive option for smaller
warehouses or less complex layouts. These technologies collectively address critical challenges in inventory management,
such as accuracy, real-time tracking, and scalability. Their integration improves operational efficiency and automates
traditionally manual processes, reducing errors and optimizing resource utilization. As inventory management is the
foundation for other warehouse operations, the strategic deployment of IPS technologies in this domain ensures broader
improvements in overall warehouse productivity.
3.4 RQ3: How can different IPS technologies for logistics management applications in a warehouse
environment be evaluated?
From the analysis in the preceding section, it is understood that IPS systems in warehouse environments feature a
wide range of technologies. Classical positioning techniques are continuously evolving, and new technologies are
emerging. Establishing an evaluation framework for IPS systems in warehouse environments would assist significantly
in comprehending the development in this area. It would provide substantial assistance to engineers engaged in related
projects. This section aims to summarize key metrics for various indoor positioning techniques and propose a
comprehensive and comparable evaluation framework for applying indoor positioning technology based on these metrics.
Unlike existing frameworks, which often generalize across warehouse tasks, this framework emphasizes criteria directly
tied to inventory management, such as real-time accuracy, adaptability to dynamic environments, and cost-effectiveness.
A review of five papers on indoor positioning technology [54] [127], [128], [129] underscores the significance of
classification methods and evaluation metrics for establishing IPS systems. An assessment system that considers energy
efficiency, cost, availability, reception range delay, scalability, and tracking accuracy was proposed by Zafari et al. [54].
Seco et al. [127] classified indoor positioning systems into four categories: geometry-based methods, minimization of the
cost function, fingerprint localization, and Bayesian techniques. A division of positioning systems into 13 categories
based on technical characteristics was systematically undertaken by Mautz et al. [128]. Detailed summaries of two review
papers proposing evaluation frameworks for indoor positioning systems have been compiled. Performance benchmarking
for indoor wireless location systems, including accuracy, precision, complexity, scalability, robustness, and cost, was
provided by Liu et al. [130]. In Liu's paper, evaluations of the IPS systems were conducted using specific data and metrics.
Zafari et al. [54] identified several key challenges in indoor localization and incorporated them as the main indicators in
their evaluation framework. While researchers have proposed various classification criteria or evaluation frameworks for
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10372
indoor positioning technologies, many are designed for general academic research. In contrast, this study focuses on
comprehensive evaluation criteria tailored explicitly for the warehouse environment.
Table 11. IPS application evaluation metrics
Metrics
Description
Evaluation method
Applicability
Using easily accessible technology that does not
require specialized hardware on the user's end is
crucial for widespread adoption.
The amount of equipment and effort needed for
deploying different indoor positioning
technologies is compared, and applicability is
evaluated as low, medium, or high.
Accuracy
The most critical aspect of a positioning system is
the accuracy with which the user/device position is
obtained. In the papers surveyed, the focus is
predominantly on localization accuracy.
Consideration for other functions or parameters is
only given if accuracy is satisfactory.
Data evaluation
Cost
While most positioning systems surveyed are in
laboratory environments and rarely mention the
cost. In this study, it is believed that high-precision
positioning significantly increases system cost.
Widespread consumer market adoption requires
reasonable cost control.
Cost assessment is based on a uniform
applicable area within the same indoor
positioning scenario. Market economic factors
are considered mainly
, and only the official
selling price of the main equipment is used for
relative evaluation as low, medium, or high.
Energy
Efficiency
The design of a positioning system must prioritize
power efficiency to enable extended operational
periods without draining the device's battery.
Therefore, reducing
energy consumption is
essential for the system's prolonged stability and
reliability.
The sum of the power consumption of major
equipment is considered. Evaluations are made
based on low, medium, or high energy
efficiency.
Scalability
The system's scalability and applicability are
considered essential.
Uncertain factors that might arise during the use
of the technology and their consequences are
considered in the evaluation, which is measured
as low, medium, or high.
In general, scholars have undertaken substantial discussions on how to evaluate IPS systems. ISO/IEC 18305:2016
International Standard, which identifies appropriate performance metrics and test & evaluation scenarios, also provides
guidance on the best ways to present and visualize T&E results. However, a critical review of this standard was conducted
by Potorti et al. [131], who believe many indicators are unsuitable for direct user applications. A perspective that IPS
must be low-cost, low-power, and require a minimal amount of new infrastructure was expressed by Wirola et al. [132].
Performance metrics proposed in the reviewed articles have been synthesized, and popular research areas indicated in
Figure 6 have been considered. Metrics of limited evaluation value in the warehouse environment, such as "robustness,"
have been excluded and merged into the "scalability" metric. Primary indicators more suitable for evaluating IPS in the
warehouse environment, including "applicability," "accuracy," "cost," "energy efficiency," and "scalability," have been
selected. The evaluation methods for these metrics are summarized in Table 11.
To comprehensively evaluate IPS technologies for inventory management, this study compares RFID, UWB, LiDAR,
IMU, Wi-Fi, VLC and Computer Vision across several critical dimensions: precision, cost, scalability, and environmental
adaptability. These dimensions are essential for ensuring that selected technologies align with the specific requirements
of inventory management tasks. UWB provides the highest precision, with localization errors typically below 10 cm,
making it ideal for dynamic inventory tasks in large or multi-level warehouses. LiDAR offers comparable precision in
static or semi-static environments, particularly high-density storage spaces. RFID and Wi-Fi, while less precise, are
effective for routine stocktaking and bulk inventory identification tasks. RFID and Wi-Fi are the most cost-effective
options, suitable for budget-constrained implementations or smaller warehouses. UWB and LiDAR, although more
expensive, justify their higher costs with superior performance in high-complexity scenarios where precision and
adaptability are paramount. Its high cost and environmental sensitivity limit LiDAR's scalability, while Wi-Fi scalability
depends on robust network infrastructure. This comparative analysis highlights that no single IPS technology universally
outperforms others; instead, its effectiveness depends on the specific priorities and constraints of the inventory
management task. For instance, UWB is best suited for high-precision dynamic tracking, whereas RFID offers an optimal
solution for large-scale, cost-sensitive implementations. These trade-offs underscore the importance of selecting
technologies based on task-specific needs, as shown in Tables 12 and 13. In summary, comprehensive data analysis and
compilation based on 38 selected articles are presented, addressing the RQs, providing substantial data support for the
subsequent discussions, and providing insights into the field of IPS in warehouse environments.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10373
Table 12. Comparison of IPS technologies in warehouse environments
Tech.
Evaluation Framework
References
Applicability
Accuracy
Cost
Energy Efficiency
Scalability
Computer Vision
High
High
Medium
Low
High
[23],[37],[64]
[50],[61],[25]
[49],[30],[44]
[45],[35],[33]
LiDAR
Medium
High
High
Medium
Medium
[23],[64],[24]
[61],[25],[52]
[41],[51],[42]
[65],[45],[46]
RFID
Medium
Low-Medium
Low
Low
Low
[39],[27],[29]
[63],[31],[48]
[32],[47],[34]
UWB
Low
High
Medium
Low
Medium
[38],[26],[59]
[60]
Wi-Fi
Low
Medium-High
Medium
Medium
Medium
[58],[28],[57]
IMU
Medium
Medium
Low
Low
Low
[61],[42]
VLC
Medium
Low
Low
Low
Low
[40],[43]
Table 13. Advantages and limitations of IPS technologies
Tech.
Advantages
Limitations
Computer
Vision
Current vision SLAM technologies offer higher
accuracy
. Less reliance on external environment
modifications is needed for positioning. Scalability in
more complex and dynamic environments is provided.
Object detection and tracking capabilities are present.
Computational requirements are high, and much
arithmetic support is needed. Positioning accuracy
may be affected by external lighting conditions
when using
ordinary monocular or binocular
cameras. Coverage is limited, generally within 10
meters, and line-of-sight is required.
LiDAR
Distance to the target point is measured more accurately
using TOF technology. A larger measurement distance
and coverage area are provided. Strong adaptability in
different lighting conditions is observed.
High costs. Object detection on transparent or
reflective surfaces may lack accuracy. Large
amounts of point cloud data require processing. A
direct line-of-sight to objects is needed for sensing.
RFID
NLOS tracking is enabled and suitable for complex and
obstructed environments like warehouses. Costs are
relatively low, especially for warehouse deployments.
Positioning accuracy is limited. The detection
range for RFID tags by the reader is limited.
Performance may be affected by environmental
factors such as metal surfaces, liquids, and
electromagnetic interference.
UWB
High-precision positioning within a few centimetres is
achieved. UWB signals can penetrate obstacles due to
high adaptability to complex environments
. Power
consumption is relatively low.
The range is limited. Initial setup may require
careful planning and calibration, increasing the up-
front workload. High initial costs may be incurred.
Wi-Fi
Wide coverage is typically provided in various indoor
environments. Existing Wi-Fi infrastructure is utilized,
making it cost-effective. Large areas can be covered, and
signals can penetrate obstacles.
Interference and signal variations may occur in
environments with high device density. Signal
transmission faces challenges such as attenuation,
reflections, and multipath interference. Reliance on
the availability and coverage of Wi-Fi access points
exists.
IMU
Real-time motion tracking and orientation information
for continuous positioning updates are provided.
Independence from external infrastructure is achieved.
High-frequency data updates are offered to track fast-
moving objects.
Measurement errors and drift over time result in
accumulated positioning errors
. Only relative
positioning information is provided. Sensitivity to
external factors like magnetic fields, temperature
variations, and vibrations is high.
VLC
Existing indoor lighting infrastructure can be utilized,
eliminating the need for specialized hardware. Operation
in the visible spectrum results in less interference. High
data rates for real-time positioning updates are enabled.
Direct line-of-sight communication between the
source and receiver is required. Performance may
be affected by ambient light conditions.
Dependency on the availability and proper
functioning of light sources exists.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10374
3.5 Challenges of IPS in Warehouse Environment
Though some intelligent systems in warehouse environments that utilize IPS have transitioned from experimental to
commercial stages, further analysis and discussion are still needed to address the proposed research questions and reveal
the challenges of IPS application in warehouse settings.
3.5.1 Complexity of environment
The indoor warehouse environment is characterized by its dynamic and complex layout, including obstacles,
fluctuating light conditions, and electronic interference [50]. Various interference sources typically plague indoor
environments, such as illuminance affecting optical sensors and temperature and sound affecting ultrasonic sensors [133].
Furthermore, the density of shelf space can compromise the wireless and optical signals, affecting positioning accuracy
[78] [134]. These complexities contribute to the unique challenges when implementing indoor positioning technologies
in warehouse environments.
3.5.2 Unknown environment
Current positioning technologies often rely on prior environmental information, as is the case with Wi-Fi [14], [135]
[136], UWB [84], and RFID [137], [138]. Pre-set beacons or signal base stations are commonly used in these technologies
[33]. However, obtaining such environmental information in real-world warehouse settings can be challenging, as can
mitigating interference with wireless base stations or accounting for randomly changing layouts. Therefore, the realization
of environment-independent positioning technology remains a challenge.
3.5.3 Balancing IPS accuracy and cost
Economic efficiency serves as a key consideration in commercial applications [24]. Higher IPS accuracy often comes
with a higher price tag [25], necessitating the development of low-cost, high-accuracy solutions. The focus has thus shifted
to balancing cost and accuracy to facilitate the broader adoption of IPS technology.
3.5.4 Multi-technology integration
Due to varying positioning principles and methods, different technologies are employed for indoor positioning, each
with its impact on indoor applications. Various factors such as accuracy, cost, and deployment difficulty often necessitate
combining multiple technologies. For example, Kwon et al. [25] utilized a blend of images, LiDAR, and IMU information
to obtain UAV attitude information for warehouse inventory applications. Challenges such as inconsistent signal
measurement units, sampling frequencies, and accuracy restrict the growth prospects of IPS technology.
3.5.5 Limited computing resources on mobile terminals
The involvement of mobile terminals in IPS is crucial, and hardware limitations can restrict the operational lifespan
and the capability to run complex positioning algorithms [24]. For instance, achieving high-speed positioning requires
substantial onboard UAV computing power, which demands more energy supplies. This limitation hampers the broader
use of IPS.
4. CONCLUSIONS
In the context of Industry 4.0, a growing need for intelligent warehouse management solutions is observed, and
attention is increasingly directed towards IPS as the foundational technology for automating and informatizing the
warehousing and logistics industry. The Scoping Review method was followed in this study, guided by the PRISMA
checklist, to identify and synthesize existing literature on IPS applications in warehouse environments from databases
within WOS, IEEE, and SCOPUS. Comprehensive survey and research results are provided in three aspects: (1) the
current state of IPS adoption in warehouse environments; (2) the technologies utilized for IPS adoption in these
environments; and (3) a framework for evaluating IPS in warehouse settings. Challenges identified in this scoping review
for the holistic application of IPS in warehouse environments, particularly inventory management.
Research Question 1 (RQ1) analysis found that IPS primarily focuses on inventory and logistics management tasks in
warehouse environments. These applications aim to upgrade existing warehouse operations to enhance efficiency and
minimize manual labor. Additionally, multiple studies aimed at integrating IPS with efforts to digitize warehouses were
identified. These kinds of applications, distinct from the previous categories, may serve as foundational steps towards
realising Industry 4.0 by enabling comprehensive digitization of warehouse processes. For Research Question 2 (RQ2), a
comparative analysis of IPS technologies revealed their respective strengths and limitations in inventory management
tasks. RFID is widely recognized for its cost-effectiveness and scalability, making it ideal for bulk inventory tracking and
routine stocktaking. UWB excels in high-precision dynamic inventory tracking in large or multi-level warehouses, while
LiDAR offers advanced 3D mapping capabilities for static or semi-static inventory spaces. Wi-Fi provides a cost-sensitive
option for smaller warehouses with simpler layouts. The findings underscore the importance of selecting IPS technologies
based on the specific requirements of inventory management tasks, such as applicability, accuracy, cost, energy efficiency
and scalability. Regarding Research Question 3 (RQ3), the study proposed a framework for evaluating IPS technologies
in inventory management, incorporating quantitative and qualitative criteria. Quantitative metrics, such as accuracy and
real-time capability, can be directly obtained from existing studies, while qualitative criteria, such as scalability and
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10375
adaptability, require interpretative analyses. The framework emphasizes the need for task-specific evaluation to address
the diverse concerns of inventory management in dynamic warehouse environments.
Future research explores integrating IPS technologies with emerging advancements in artificial intelligence and deep
learning. For instance, combining IPS data with deep learning algorithms can enhance localization accuracy and predictive
capabilities in dynamic warehouse environments. Additionally, developing hybrid IPS solutions, leveraging the
complementary strengths of technologies such as UWB and LiDAR presents a promising avenue for addressing the trade-
offs between precision, scalability, and cost. Finally, applying IPS in complex, multi-modal logistics systems and highly
automated warehouses offers significant innovation potential, paving the way for realizing Industry 4.0.
ACKNOWLEDGEMENTS
The authors thank SEGi University, Malaysia, for providing the laboratory facilities necessary for this research. The
authors also thank the anonymous reviewers and editors for their constructive suggestions, which significantly improved
this manuscript. This research received no funding from public, private, or non-profit funding agencies.
CONFLICT OF INTEREST
The authors declare no conflicts of interest, either financial or non-financial, including political, personal, or professional
relationships that could have influenced the manuscript.
AUTHORS CONTRIBUTION
Xiaodong Zhang (Writing – original draft; Conceptualization; Methodology; Validation; Formal analysis; Data curation;
Visualization)
Yong Chai Tan (Writing – review & editing; Project administration; Supervision)
Vin Cent Tai (Writing – review & editing; Resources)
Yanan Hao (Validation; Investigation)
AVAILABILITY OF DATA AND MATERIALS
The data supporting this study’s findings are available on request from the corresponding author.
ETHICS STATEMENT
Not applicable
REFERENCES
[1] S. Zimmermann, R. Poranne, S. Coros, “Go fetch! - Dynamic grasps using Boston Dynamics Spot with external
robotic arm,” in IEEE International Conference on Robotics and Automation, vol. 1, pp. 4488–4494, 2021.
[2] M. Javaid, A. Haleem, R. P. Singh, R. Suman, “Substantial capabilities of robotics in enhancing industry 4.0
implementation,” Cognitive Robotics, vol. 1, pp. 58–75, 2021.
[3] The New York Times, “Robots aren’t done reshaping warehouses,” New York, USA [Online], 2023. Available:
www.nytimes.com
[4] J. M. Davila Delgado, L. Oyedele, A. Ajayi, L. Akanbi, O. Akinade, M. Bilal, et al., “Robotics and automated
systems in construction: Understanding industry-specific challenges for adoption,” Journal of Building
Engineering, vol. 26, p. 100868, 2019.
[5] J. Holland, L. Kingston, C. McCarthy, E. Armstrong, P. O’Dwyer, F. Merz, et al., “Service robots in the healthcare
sector,” Robotics, vol. 10, no. 1, pp. 1–47, 2021.
[6] G. Ya n g , Z. Pang, M. J. Deen, M. Dong, Y. Zhang, N. Lovell, et al., “Homecare robotic systems for healthcare
4.0: visions and enabling technologies,” IEEE Journal of Biomedical and Health Informatics, vol. 24, no. 9, pp.
2535–2549, 2020.
[7] N. Li, L. Guan, Y. Gao, S. Du, M. Wu, X. Guang, et al., “Indoor and outdoor low-cost seamless integrated
navigation system based on the integration of INS/GNSS/LIDAR system,” Remote Sensing, vol. 12, no. 19, p.
3271, 2020.
[8] E. D. Kaplan, C. Hegarty. Understanding GPS/GNSS: Principles and Applications. 3rd Eds. United States: Artech
House, 2017.
[9] X. Zhu, W. Qu, T. Qiu, L. Zhao, M. Atiquzzaman, D. O. Wu, “Indoor intelligent fingerprint-based localization:
principles, approaches and challenges,” IEEE Communications Surveys & Tutorials, vol. 22, no. 4, pp. 2634–
2657, 2020.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10376
[10] L. Zhen, H. Li, “A literature review of smart warehouse operations management,” Frontiers of
Engineering Management, vol. 9, no. 1, pp. 31–55, 2022.
[11] T. M. Fernández-Caramés, O. Blanco-Novoa, I. Froiz-Míguez, P. Fraga-Lamas, “Towards an autonomous industry
4.0 warehouse: A UAV and blockchain-based system for inventory and traceability applications in big data-driven
supply chain management,” Sensors, vol. 19, no. 10, p. 2394, 2019.
[12] S. Janasekaran, S. Mann, “A case study: implementation of automated guided vehicle replacing trollies
transportation in manufacturing industry to reduce motion waste in lean manufacturing practicing factory,”
Journal of Engineering & Technological Advances, vol. 5, no. 1, pp. 1–7, 2020.
[13] I. Kalinov, A. Petrovsky, V. Ilin, E. Pristanskiy, M. Kurenkov, V. Ramzhaev, “WareVision: CNN barcode
detection-based UAV trajectory optimization for autonomous warehouse stocktaking,” IEEE Robotics and
Automation Letters, vol. 5, no. 4, pp. 6647–6653, 2020.
[14] P. Octaviani, W. Ce, “Inventory placement mapping using bluetooth low energy beacon technology for
warehouses,” in International Conference on Information Management and Technology (ICIMTech), vol. 1, pp.
354–359, 2020.
[15] M. Beul, N. Krombach, M. Nieuwenhuisen, D. Droeschel, S. Behnke, “Autonomous navigation in a warehouse
with a cognitive micro aerial vehicle,” in Robot Operating System (ROS): The Complete Reference, vol. 2, pp.
487–524, 2017.
[16] H. Arksey, L. O’Malley, “Scoping studies: towards a methodological framework,” International Journal of Social
Research Methodology, vol. 8, no. 1, pp. 19–32, 2005.
[17] A. C. Tricco, E. Lillie, W. Zarin, K. O'Brien, H. Colquhoun, D. Levac, et al., “PRISMA extension for scoping
reviews (PRISMA-ScR): Checklist and explanation,” Annals of Internal Medicine, vol. 169, no. 7, pp. 467–473,
2018.
[18] A. Liberati, D. G. Altman, J. Tetzlaff, C. Mulrow, P. C. Gøtzsche, P. J. Devereaux, et al., “The PRI S M A statement
for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: Explanation
and elaboration,” Annals of Internal Medicine, vol. 151, no. 4, p. 65, 2009.
[19] J. Wichmann, “Indoor positioning systems in hospitals: A scoping review,” Digital Health, vol. 8, p.
205520762210816, 2022.
[20] V. Bellavista-Parent, J. Torres-Sospedra, A. Perez-Navarro, “New trends in indoor positioning based on WiFi and
machine learning: A systematic review,” in International Conference on Indoor Positioning and Indoor
Navigation, pp. 1–8, 2021.
[21] H. L. Kopsco, R. L. Smith, S. J. Halsey, “A scoping review of species distribution modeling methods for tick
vectors,” Frontiers in Ecology and Evolution, vol. 10, p. 893016, 2022.
[22] R. Sharma, S. Gulati, A. Kaur, A. Sinhababu, R. Chakravarty, “Research discovery and visualization using
ResearchRabbit: A use case of AI in libraries,” COLLNET Journal of Scientometrics and Information
Management, vol. 16, no. 2, pp. 215–237, 2022.
[23] R. Krug, T. Stoyanov, V. Tincani, H. Andreasson, R. Mosberger, G. Fantoni, “The next step in robot
commissioning: Autonomous picking and palletizing,” IEEE Robotics and Automation Letters, vol. 1, no. 1, pp.
546–553, 2016.
[24] M. Beul, D. Droeschel, M. Nieuwenhuisen, J. Quenzel, S. Houben, S. Behnke, “Fast autonomous flight in
warehouses for inventory applications,” IEEE Robotics and Automation Letters, vol. 3, no. 4, pp. 3121–3128,
2018.
[25] W. Kwon, J. H. Park, M. Lee, J. Her, S.H. Kim, J. W. Seo, “Robust autonomous navigation of unmanned aerial
vehicles for warehouses’ inventory application,” IEEE Robotics and Automation Letters, vol. 5, no. 1, pp. 243–
249, 2020.
[26] M. Khalaf-Allah, “Particle filtering for three-dimensional TDoA-based positioning using four anchor nodes,”
Sensors, vol. 20, no. 16, p. 4516, 2020.
[27] C. Li, E. Tanghe, D. Plets, P. Suanet, J. Hoebeke, E. D. Poorter, “ReLoc: Hybrid RSSI- and phase-based relative
UHF-RFID tag localization with COTS devices,” IEEE Transactions on Instrumentation and Measurement, vol.
69, no. 10, pp. 8613–8627, 2020.
[28] V. Sircoulomb, H. Chafouk, “A constrained Kalman filter for Wi-Fi-based indoor localization with flexible space
organization,” Sensors, vol. 22, no. 2, p. 428, 2022.
[29] C. Li, E. Tanghe, P. Suanet, D. Plets, J. Hoebeke, E. D. Poorter, “ReLoc 2.0: UHF-RFID relative localization for
drone-based inventory management,” IEEE Transactions on Instrumentation and Measurement, vol. 70, pp. 1–
13, 2021.
[30] E. Martinez-Martin, E. Ferrer, I. Vasilev, A. P. Del Pobil, “The UJI aerial librarian robot: A quadcopter for visual
library inventory and book localization,” Sensors, vol. 21, no. 4, p. 1079, 2021.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10377
[31] A. Alajami, G. Moreno, R. Pous, “Design of a UAV for autonomous RFID-based dynamic inventories using
stigmergy for mapless indoor environments,” Drones, vol. 6, no. 8, p. 208, 2022.
[32] C. Wu, Z. Gong, B. Tao, K. Tan, Z. Gu, Z. Yin, “Rf-slam: UHF-RFID based simultaneous tags mapping and robot
localization algorithm for smart warehouse position service,” IEEE Transactions on Industrial Informatics, vol.
19, pp. 1–11, 2023.
[33] M. Ekici, A. Ç. Seçkin, A. Özek, C. Karpuz, “Warehouse drone: indoor positioning and product counter with
virtual fiducial markers,” Drones, vol. 7, no. 1, p. 3, 2023.
[34] M. Gareis, M. Hehn, P. Stief, G. Körner, C. Birkenhauer, J. Trabert, “Novel UHF-RFID listener hardware
architecture and system concept for a mobile robot based MIMO SAR RFID localization,” IEEE Access, vol. 9,
pp. 497–510, 2021.
[35] J. Martinez-Carranza, L. O. Rojas-Perez, “Warehouse inspection with an autonomous micro air vehicle,”
Unmanned Systems, vol. 10, no. 4, pp. 329–342, 2022.
[36] J. Stanko, F. Stec, J. Rodina, “Process automation of warehouse inspection using an autonomous unmanned aerial
vehicle,” MM Science Journal, vol. 1, pp. 5864–5869, 2022.
[37] M. Himstedt, E. Maehle, “Online semantic mapping of logistic environments using RGB-D cameras,”
International Journal of Advanced Robotic Systems, vol. 14, no. 4, pp. 1-13, 2017.
[38] D. Shi, H. Mi, E. Collins, J. Wu, “An indoor low-cost and high-accuracy localization approach for AGVs,” IEEE
Access, vol. 8, pp. 50085–50090, 2020.
[39] J. H. Teo, A. Loganathan, P. Goh, N. S. Ahmad, “Autonomous mobile robot navigation via RFID signal strength
sensing,” International Journal of Mechanical Engineering and Robotics Research, vol. 8, pp. 1140–1144, 2020.
[40] P. Louro, M. Vieira, M. Vieira, “Bidirectional visible light communication,” Optical Engineering, vol. 59, no. 12,
p. 127109, 2020.
[41] R. Lin, H. Huang, M. Li, “An automated guided logistics robot for pallet transportation,” Assembly Automation,
vol. 41, no. 1, pp. 45–54, 2021.
[42] H. Zhang, L. Yu, S. Fei, “Design of dual-LiDAR high precision natural navigation system,” IEEE Sensors Journal,
vol. 22, no. 7, pp. 7231–7239, 2022.
[43] P. Louro, M. Vieira, M. A. Vieira, “Geolocalization and navigation by visible light communication to address
automated logistics control,” Optical Engineering, vol. 61, no. 1, p. 016104, 2022.
[44] Q. Yang, Y. Lian, Y. Liu, W. Xie, Y. Yang, “Multi-Agv tracking system based on global vision and apriltag in
smart warehouse,” Journal of Intelligent & Robotic Systems, vol. 104, no. 3, 2022.
[45] Y. Zhang, Y. Zhou, H. Li, H. Hao, W. Chen, W. Zhan, “The navigation system of a logistics inspection robot based
on multi-sensor fusion in a complex storage environment,” Sensors, vol. 22, no. 20, p. 7794, 2022.
[46] T. Kim, H. Jeon, D. Lee, “A multi-layered 3d NDT scan-matching method for robust localization in logistics
warehouse environments,” Sensors, vol. 23, no. 5, p. 2671, 2023.
[47] L. Mo, C. Li, “Passive UHF-RFID localization based on the similarity measurement of virtual reference tags,”
IEEE Transactions on Instrumentation and Measurement, vol. 68, no. 8, pp. 2926–2933, 2018.
[48] P. Tripicchio, S. D’Avella, M. Unetti, “Efficient localization in warehouse logistics: a comparison of LMS
approaches for 3D multilateration of passive UHF RFID tags,” The International Journal of Advanced
Manufacturing Technology, vol. 120, no. 7–8, pp. 4977–4988, 2022.
[49] K. Prakash, M. N. G. Mohamed, S. Chakravorty, Z. Hasnain, “Structure aided odometry (SAO): A novel analytical
odometry technique based on semi-absolute localization for precision-warehouse robotic assistance in
environments with low feature variation,” Journal of Intelligent & Robotic Systems, vol. 102, p. 72, 2021.
[50] K. Mohta, M. Watterson, Y. Mulgaonkar, S. Liu, C. Qu, A. Makineni, et al., “Fast, autonomous flight in GPS-
denied and cluttered environments,” Journal of Field Robotics, vol. 35, no. 1, pp. 101–120, 2018.
[51] H. Wang, C. Wang, L. Xie, “Intensity-slam: Intensity assisted localization and mapping for large scale
environment,” IEEE Robotics and Automation Letters, vol. 6, no. 2, pp. 1715–1721, 2021.
[52] R. M. Gago, M. Y. A. Pereira, G. A. S. Pereira, “An aerial robotic system for inventory of stockpile warehouses,”
Engineering Reports, vol. 3, no. 9, p. 12396, 2021.
[53] A. Basiri, E. S. Lohan, T. Moore, A. Winstanley, P. Peltola, C. Hill, et al., “Indoor location-based services
challenges, requirements and usability of current solutions,” Computer Science Review, vol. 24, pp. 1–12, 2017.
[54] F. Zafari, A. Gkelias, K. K. Leung, “A survey of indoor localization systems and technologies,” IEEE
Communications Surveys & Tutorials, vol. 21, no. 3, pp. 2568–2599, 2019.
[55] G. M. Mendoza-Silva, J. Torres-Sospedra, J. Huerta, “A meta-review of indoor positioning systems,” Sensors,
vol. 19, no. 20, p. 4507, 2019.
[56] X. Xu, F. Pang, Y. Ran, Y. Bai, L. Zhang, Z. Tan, “An indoor mobile robot positioning algorithm based on adaptive
federated Kalman filter,” IEEE Sensors Journal, vol. 21, no. 20, pp. 23098-23107, 2021.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10378
[57] F. Xiao, S. Zhang, S. Tang, S. Shen, H. Dong, Y. Zhong, “Wision: Bolstering MAV 3d indoor state estimation by
embracing multipath of wifi,” IEEE Transactions on Vehicular Technology, vol. 72, no. 1, pp. 253–266, 2023.
[58] A. Vashist, M. Patrick Li, A. Ganguly, S. Manoj, C. Hochgraf, R. Ptucha, “KF-Loc: A Kalman filter and machine
learning integrated localization system using consumer-grade millimeter-wave hardware,” IEEE Consumer
Electronics Magazine, vol. 11, no. 4, pp. 65–77, 2022.
[59] W. Zhao, J. Panerati, A. P. Schoellig, “Learning-based bias correction for time difference of arrival ultra-wideband
localization of resource-constrained mobile robots,” IEEE Robotics and Automation Letters, vol. 6, no. 2, pp.
3639–3646, 2021.
[60] S. Monica, G. Ferrari, “Robust UWB-based localization with application to automated guided vehicles,”
Advanced Intelligent Systems, vol. 3, no. 4, p. 2000083, 2021.
[61] D. Kim, S. Shin, and I. S. Kweon, “On-line initialization and extrinsic calibration of an inertial navigation system
with a relative preintegration method on manifold,” IEEE Transactions on Automation Science and Engineering,
vol. 15, no. 3, pp. 1272–1285, 2018.
[62] W. Kwon, J. H. Park, M. Lee, J. Her, S.-H. Kim, J.-W. Seo, “Robust autonomous navigation of unmanned aerial
vehicles (UAVs) for warehouses' inventory application,” IEEE Robotics and Automation Letters, vol. 5, no. 1, pp.
243–249, 2020.
[63] F. Shamsfakhr, A. Motroni, L. Palopoli, A. Buffi, P. Nepa, D. Fontanelli, “Robot localisation using UHF-RFID
tags: A Kalman smoother approach,” Sensors, vol. 21, no. 3, p. 717, 2021.
[64] M. Jung, J.-B. Song, “Robust mapping and localization in indoor environments,” Intelligent Service Robotics,
vol. 10, no. 1, pp. 55–66, 2017.
[65] M. Chang, “Real-time multi-fusion perceptron architecture for autonomous drones,” Journal of the Chinese
Institute of Engineers, vol. 45, no. 7, pp. 621–631, 2022.
[66] B. Wang, X. Liu, B. Yu, R. Jia, X. Gan, “An improved WiFi positioning method based on fingerprint clustering
and signal weighted Euclidean distance,” Sensors, vol. 19, no. 10, p. 2300, 2019.
[67] X. Li, Z. D. Deng, L. T. Rauchenstein, T. J. Carlson, “Contributed Review: Source-localization algorithms and
applications using time of arrival and time difference of arrival measurements,” Review of Scientific Instruments,
vol. 87, no. 4, p. 041502, 2016.
[68] C. Yan g , H. Shao, “WiFi-based indoor positioning,” IEEE Communications Magazine, vol. 53, no. 3, pp. 150–
157, 2015.
[69] P. Wu, S. Su, Z. Zuo, X. Guo, B. Sun, X. Wen, “Time difference of arrival (TDoA) localization combining
weighted least squares and firefly algorithm,” Sensors, vol. 19, no. 11, p. 2554, 2019.
[70] S.-H. Yang, H.-S. Kim, Y.-H. Son, S.-K. Han, “Three-dimensional visible light indoor localization using AOA
and RSS with multiple optical receivers,” Journal of Lightwave Technology, vol. 32, no. 14, pp. 2480–2485, 2014.
[71] M. Kotaru, K. Joshi, D. Bharadia, S. Katti, “Spotfi: Decimeter level localization using wifi,” in Pro ceedings of
the ACM Conference on Special Interest Group on Data Communication, pp. 269–282, 2015.
[72] Y. C he ng , T. Zhou, “UWB indoor positioning algorithm based on TDOA technology,” in International conference
on information technology in medicine and education, pp. 777–782, 2019.
[73] W. Gong, J. Liu, “RoArray: Towards more robust indoor localization using sparse recovery with commodity
WiFi,” IEEE Transactions on Mobile Computing, vol. 18, no. 6, pp. 1380–1392, 2018.
[74] S. Shi, S. Sigg, Y. Ji, “Probabilistic fingerprinting based passive device-free localization from channel state
information,” in IEEE 83rd Vehicular Technology Conference, pp. 1–5, 2016.
[75] M. T. Hoang, B. Yuen, X. Dong, T. Lu, R. Westendorp, K. Reddy, “Recurrent neural networks for accurate RSSI
indoor localization,” IEEE Internet of Things Journal, vol. 6, no. 6, pp. 10639–10651, 2019.
[76] A. Brunello, A. Montanari, N. Saccomanno, “Towards interpretability in fingerprint based indoor positioning:
May attention be with us,” Expert Systems with Applications, vol. 231, p. 120679, 2023.
[77] Y. Ma, N. Selby, F. Adib, “Minding the billions: Ultra-wideband localization for deployed RFID tags,” in
Proceedings of the 23rd Annual International Conference on Mobile Computing and Networking, vol. 6, pp. 248–
260, 2017.
[78] J. Tiemann, Y. Elmasry, L. Koring, C. Wietfeld, “ATLAS FaST: Fast and simple scheduled TDOA for reliable
ultra-wideband localization,” in International Conference on Robotics and Automation, pp. 2554–2560, 2019.
[79] S. Marano, W. M. Gifford, H. Wymeersch, M. Z. Win, “NLOS identification and mitigation for localization based
on UWB experimental data,” IEEE Journal on Selected Areas in Communications, vol. 28, no. 7, pp. 1026–1035,
2010.
[80] G. Zhang, Z. Deng, L. Wen, L. Ge, H. Ke, J. Jiao, “An UWB location algorithm for indoor NLOS environment,”
in Ubiquitous Positioning, Indoor Navigation and Location-Based Services, pp. 1–6, 2018.
[81] K. Yu, K. Wen, Y. Li, S. Zhang, K. Zhang, “A novel NLOS mitigation algorithm for UWB localization in harsh
indoor environments,” IEEE Transactions on Vehicular Technology, vol. 68, no. 1, pp. 686–699, 2018.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10379
[82] S. Go, S. Kim, J.-W. Chong, “An efficient non-line-of-sight error mitigation method for TOA measurement in
indoor environments,” in International Conference on Ubiquitous Information Management and Communication:
Proceedings, vol. 1, pp. 1–5, 2014.
[83] J. Wang, M. Wang, D. Yang, F. Liu, Z. Wen, “UWB positioning algorithm and accuracy evaluation for different
indoor scenes,” International Journal of Image and Data Fusion, vol. 12, no. 3, pp. 203–225, 2021.
[84] N. Macoir, J. Bauwens, B. Jooris, B. V. Herbruggen, J. Rossey, J. Hoebeke, E. D. Poorter, “UWB localization
with battery-powered wireless backbone for drone-based inventory management,” Sensors, vol. 19, no. 3, p. 467,
2019.
[85] W. Vinicchayakul, S. Promwong, P. Supanakoon, “Study of UWB indoor localization using fingerprinting
technique with different number of antennas,” in International Computer Science and Engineering Conference,
vol. 1, pp. 1–4, 2016.
[86] S. Djosic, I. Stojanovic, M. Jovanovic, T. Nikolic, G. L. Djordjevic, “Fingerprinting-assisted UWB-based
localization technique for complex indoor environments,” Expert Systems with Applications, vol. 167, p. 114188,
2021.
[87] J. Tiemann, F. Schweikowski, C. Wietfeld, “Design of an UWB indoor-positioning system for UAV navigation in
GNSS-denied environments,” in International Conference on Indoor Positioning and Indoor Navigation (IPIN),
vol. 1, pp. 1–7, 2015.
[88] C. Li, L. Mo, D. Zhang, “Review on UHF RFID localization methods,” IEEE Journal of Radio Frequency
Identification, vol. 3, no. 4, pp. 205–215, 2019.
[89] J. Hightower, R. Want, G. Borriello, “SpotON: An indoor 3D location sensing technology based on RF signal
strength,” Technical Report, University of Washington, no. 2000-02-02, 2000.
[90] F. Gu, S. Valaee, K. Khoshelham, J. Shang, R. Zhang, “Landmark graph-based indoor localization,” IEEE Internet
of Things Journal, vol. 7, no. 9, pp. 8343–8355, 2020.
[91] B. Huang, J. Liu, W. Sun, F. Yang, “A robust indoor positioning method based on Bluetooth low energy with
separate channel information,” Sensors, vol. 19, no. 16, p. 3487, 2019.
[92] B. Hu, H. Peng, Z. Sun, “LANDMARC localization algorithm based on weight optimization,” Chinese Journal
of Electronics, vol. 27, no. 6, pp. 1291–1296, 2018.
[93] H. Zhou-guo, L. Fang, Y. Yi, “An improved indoor UHF RFID localization method based on deviation correction,”
in 4th International Conference on Information Science and Control Engineering, pp. 1401–1404, 2017.
[94] S. Subedi, E. Pauls, Y. D. Zhang, “Accurate localization and tracking of a passive RFID reader based on RSSI
measurements,” IEEE Journal of Radio Frequency Identification, vol. 1, no. 2, pp. 144–154, 2017.
[95] Y. Li, H. Xu, and P. Li, “RFID-Based WIMEC-LANDMARC indoor location algorithm,” in International
Conferences on Internet of Things, vol. 1, pp. 448–455. 2020.
[96] K. Panta, J. Armstrong, “Indoor localisation using white LEDs,” Electronics Letters, vol. 48, no. 4, pp. 228–230,
2012.
[97] H. Li, H. Huang, Y. Xu, Z. Wei, S. Yuan, P. Lin, “A fast and high-accuracy real-time visible light positioning
system based on single LED lamp with a beacon,” IEEE Photonics Journal, vol. 12, no. 6, pp. 1–12, 2020.
[98] M. Ibrahim, O. Moselhi, “Inertial measurement unit based indoor localization for construction applications,”
Automation in construction, vol. 71, pp. 13–20, 2016.
[99] W. A. Gill, I. Howard, I. Mazhar, K. McKee, “A review of MEMS vibrating gyroscopes and their reliability issues
in harsh environments,” Sensors, vol. 22, no. 19, p. 7405, 2022.
[100] Y. Wu, H. B. Zhu, Q. X. Du, S. M. Tang, “A survey of the research status of pedestrian dead reckoning systems
based on inertial sensors,” International Journal of Automation and Computing, vol. 16, no. 1, pp. 65–83, 2019.
[101] X. Tong, Y Su, Z Li, C Si, G Han, J Ning, F Yang, “A double-step unscented Kalman filter and HMM-based zero-
velocity update for pedestrian dead reckoning using MEMS sensors,” IEEE Transactions on Industrial
Electronics, vol. 67, no. 1, pp. 581–591, 2019.
[102] Y. Wu, F. Tang, H. Li, “Image-based camera localization: an overview,” Visual Computing for Industry,
Biomedicine, and Art, vol. 1, no. 1, p. 8, 2018.
[103] D. J. Kriegman, E. Triendl, T. O. Binford, “Stereo vision and navigation in buildings for mobile robots,” IEEE
Transactions on Robotics and Automation, vol. 5, no. 6, pp. 792–803, 1989.
[104] M. J. Milford, G. F. Wyeth, “SeqSLAM: Visual route-based navigation for sunny summer days and stormy winter
nights,” in IEEE International Conference on Robotics and Automation, pp. 1643–1649, 2012.
[105] H. W. Ho, G. C. de Croon, Q. Chu, “Distance and velocity estimation using optical flow from a monocular
camera,” International Journal of Micro Air Vehicles, vol. 9, no. 3, pp. 198–208, 2017.
[106] R. Arroyo, P. F. Alcantarilla, L. M. Bergasa, E. Romera, “Are you able to perform a life-long visual topological
localization,” Autonomous Robots, vol. 42, pp. 665–685, 2018.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10380
[107] R. Arandjelovic, P. Gronat, A. Torii, T. Pajdla, J. Sivic, “NetVLAD: CNN Architecture for Weakly Supervised
Place Recognition,” in IEEE Conference on Computer Vision and Pattern Recognition, pp. 5297–5307, 2016.
[108] R. Arandjelovic, A. Zisserman, “All about VLAD,” in IEEE Conference on Computer Vision and Pattern
Recognition: Proceedings, pp. 1578–1585, 2013.
[109] H. Zhou, K. Ni, Q. Zhou, T. Zhang, “An SFM algorithm with good convergence that addresses outliers for
realizing mono-SLAM,” IEEE Transactions on Industrial Informatics, vol. 12, no. 2, pp. 515–523, 2016.
[110] R. Mur-Artal, J. M. M. Montiel, J. D. Tardos, “ORB-SLAM: A versatile and accurate monocular SLAM system,”
IEEE Transactions on Robotics, vol. 31, no. 5, pp. 1147–1163, 2015.
[111] R. Mur-Artal, J. D. Tardós, “Orb-slam2: An open-source SLAM system for monocular, stereo, and RGB-D
cameras,” IEEE Transactions on Robotics, vol. 33, no. 5, pp. 1255–1262, 2017.
[112] C. Campos, R. Elvira, J. J. G. Rodríguez, J. M. Montiel, J. D. Tardós, “Orb-slam3: An accurate open-source library
for visual, visual-inertial, and multimap slam,” IEEE Transactions on Robotics, vol. 37, no. 6, pp. 1874–1890,
2021.
[113] T. Qin, P. Li, S. Shen, “Vins-mono: A robust and versatile monocular visual-inertial state estimator,” IEEE
Transactions on Robotics, vol. 34, no. 4, pp. 1004–1020, 2018.
[114] L. Campos-Macías, R. Aldana-López, R. de la Guardia, J. I. Parra-Vilchis, D. Gómez-Gutiérrez, “Autonomous
navigation of MAVs in unknown cluttered environments,” Journal of Field Robotics, vol. 38, no. 2, pp. 307–326,
2021.
[115] Y. L i, J. Ibanez-Guzman, “Lidar for autonomous driving: The principles, challenges, and trends for automotive
lidar and perception systems,” IEEE Signal Processing Magazine, vol. 37, no. 4, pp. 50–61, 2020.
[116] M. U. Khan, S. A. A. Zaidi, A. Ishtiaq, S. U. R. Bukhari, S. Samer, A. Farman, “A comparative survey of lidar-
slam and lidar-based sensor technologies,” in Mohammad Ali Jinnah University International Conference on
Computing, pp. 1–8, 2021.
[117] E. Sánchez, M. Botsch, B. Huber, A. García, “High precision indoor positioning by means of LiDAR,” in DGON
Inertial Sensors and Systems Symposium, pp. 1–20, 2019.
[118] Y. -T. Wang, C.C. Peng, A. A. Ravankar, A. Ravankar, “A single LiDAR-based feature fusion indoor localization
algorithm,” Sensors, vol. 18, no. 4, p. 1294 2018.
[119] S. H. I. Pengcheng, Y. E. Qin, Z. Shaoming, D. Haifeng, “Localization initialization for multi-beam LiDAR
considering indoor scene feature,” Acta Geodaetica et Cartographica Sinica, vol. 50, no. 11, p. 1594, 2021.
[120] H. Wang, C. Wang, C. L. Chen, L. Xie, “F-LOAM : Fast LiDAR odometry and mapping,” in IEEE/RSJ
International Conference on Intelligent Robots and Systems, pp. 4390–4396, 2021.
[121] D. V. Nam, K. Gon-Woo, “Solid-state LiDAR based-SLAM: A concise review and application,” in IEEE
International Conference on Big Data and Smart Computing, pp. 302–305, 2021.
[122] F. Che, Q. Z. Ahmed, P. I. Lazaridis, P. Sureephong, T. Alade, “Indoor positioning system (IPS) using ultra-wide
bandwidth (UWB)—for industrial internet of things (iIoT),” Sensors, vol. 23, no. 12, p. 5710, 2023.
[123] Y. Tao, L. Wu, J. Sidén, G. Wang, “Monte Carlo-based indoor RFID positioning with dual-antenna joint
rectification,” Electronics, vol. 10, no. 13, p. 2764, 2021.
[124] Y. Chen, W. Guan, J. Li, H. Song, “Indoor real-time 3-D visible light positioning system using fingerprinting and
extreme learning machine,” IEEE Access, vol. 8, pp. 13875–13886, 2020.
[125] Y. Hao, V. C. Tai, and Y. C. Tan, “A Systematic Stereo Camera Calibration Strategy: Leveraging Latin Hypercube
Sampling and 2k Full-Factorial Design of Experiment Methods,” Sensors, vol. 23, no. 19, p. 8240, 2023.
[126] A. J. Davison, “Real-time simultaneous localisation and mapping with a single camera,” in IEEE International
Conference on Computer Vision: Proceedings, vol. 2, pp. 1403–1410, 2003.
[127] F. Seco, A. R. Jimenez, C. Prieto, J. Roa, K. Koutsou, “A survey of mathematical methods for indoor localization,”
in IEEE International Symposium on Intelligent Signal Processing, pp. 9–14, 2009.
[128] R. Mautz, S. Tilch, “Survey of optical indoor positioning systems,” in International Conference on Indoor
Positioning and Indoor Navigation, pp. 1–7, 2011.
[129] X. Guo, N. Ansari, F. Hu, Y. Shao, N. R. Elikplim, L. Li, “A survey on fusion-based indoor positioning,” IEEE
Communications Survey & Tutorials, vol. 22, no. 1, pp. 566–594, 2020.
[130] H. Liu, H. Darabi, P. Banerjee, J. Liu, “Survey of wireless indoor positioning techniques and systems,” IEEE
Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews), vol. 37, no. 6, pp. 1067–
1080, 2007.
[131] F. Potorti, A. Crivello, P. Barsocchi, F. Palumbo, “Evaluation of indoor localisation systems: Comments on the
ISO/IEC 18305 standard,” in International Conference on Indoor Positioning and Indoor Navigation, pp. 1–7,
2018.
X. D. Zhang et al. │ Journal of Mechanical Engineering and Sciences │ Vol. 18, Issue 4 (2024)
journal.ump.edu.my/jmes 10381
[132] L. Wirola, T. A. Laine, J. Syrjarinne, “Mass-market requirements for indoor positioning and indoor navigation,”
in International Conference on Indoor Positioning and Indoor Navigation, pp. 1–7, 2010.
[133] S. E. Li, G. Li, J. Yu, C. Liu, B. Cheng, J. Wang, et al., “Kalman filter-based tracking of moving objects using
linear ultrasonic sensor array for road vehicles,” Mechanical Systems and Signal Processing, vol. 98, pp. 173–
189, 2018.
[134] B. Xie, K. Chen, G. Tan, M. Lu, Y. Liu, J. Wu, T. He, “LIPS: A light intensity--based positioning system for indoor
environments,” ACM Transactions on Sensor Networks, vol. 12, no. 4, p. 1–27, 2016.
[135] C. Röhrig, D. Heß, C. Kirsch, F. Künemund, “Localization of an omnidirectional transport robot using IEEE
802.15.4a ranging and laser range finder,” in IEEE/RSJ International Conference on Intelligent Robots and
Systems, pp. 3798–3803, 2010.
[136] A. Va s hi s t, D. R. Bhanushali, R. Relyea, C. Hochgraf, A. Ganguly, P. D. Sai, “Indoor wireless localization using
consumer-grade 60 GHz equipment with machine learning for intelligent material handling,” in IEEE
International Conference on Consumer Electronics, pp. 1–6, 2020.
[137] G. L. Scalia, G. Aiello, M. Enea, R. Micale, “Preliminary analysis of warehouse localization systems based on
RFID technology,” International Journal of RF Technologies, vol. 2, no. 1, pp. 23–36, 2010.
[138] C. Wu, Z. Gong, B. Tao, K. Tan, Z. Gu, Z. Yin, “Rf-slam: Uhf-RFID based simultaneous tags mapping and robot
localization algorithm for smart warehouse position service,” IEEE Transactions on Industrial Informatics, vol.
19, no. 12, pp. 11765–11775, 2023.
[139] S. Cheng, S. Wang, W. Guan, H. Xu, P. Li, “3DLRA: An RFID 3D indoor localization method based on deep
learning,” Sensors, vol. 20, no. 9, p. 2731, 2020.
