Conference PaperPDF Available

Site inspection drone: A solution for inspecting and regulating construction sites

Authors:
  • Dubai Futue Labs
Site Inspection Drone: A Solution for Inspecting and
Regulating Construction Sites
Reem Ashour, Tarek Taha, Fahad Mohamed, Eman Hableel, Yasmeen Abu Kheil,
Malak Elsalamouny, Maha Kadadha, Kasturi Rangan, Jorge Dias, Lakmal Seneviratne, Guowei Cai
Khalifa University Robotics Institute, Khalifa University of Science, Technology and Research, Abu Dhabi, UAE
Abstract—Motivated by the need to regulate and regularly
inspect construction sites, we developed a real-time robust drone
system that performs site inspection and violation detection
in construction sites. The proposed system uses the drone to
inspect and monitor numerous types of construction sites remotely
from any control station or a nearby locations. Our drone
system was developed as an entry to participate in ”Drones
For Good” 2015 competition, and qualified to the finals. Our
proposed technological solution presented in this paper introduces
the possible utilization of drone technology in the construction
industry. We will first introduce the motivation and benefits of
using this system, and how it’s usage will increase the ability
to control the health and safety, efficiency, and cost of the site
inspection process. We will then detail the technical components of
the drone, and present the results of experiments and evaluations
conducted in realistic scenarios.
I. INT ROD UC TI ON
United Arab Emirates is one of the most developed and
competitive countries in the construction business in the world.
While aiming to maintain sustainable development, the boom
in the construction sector across the country is increasing
the challenges faced by local municipalities to control, reg-
ulate and monitor these construction sites. These challenges
are mainly concerned with Preventing Labor force abuse,
Monitoring health and safety standards of construction sites,
and Stopping unlicensed constructions. Therefore, utilizing
and implementing new technologies such as ”Drones” can be
useful in this sector to: insure safety of workers, improve time
efficiency; reduce paper work; and automate the entire process.
Various applications of using drones to inspect different
structures were proposed in the literature. In [1], Secutronic
INSPIRE drones were used by Nokia to inspect towers, test
line of sight and plan radio sites in order to optimize the
telecommunication networks. Drones were also used by San
Diego Gas and Electric Company to inspect power lines and
pipelines [2]. In this case, drones will fly over difficult to
reach areas and stream images that show the condition of
power lines and pipelines. Boeing also investigated the usage
of drone platforms to inspect aircraft for lightning scratches
and burn [3].
To assess the benefits of utilizing drones for site in-
spections, five different criteria were selected to evaluate the
benefits of our drone solution and compare it to the current
existing methods. Each criterion is then weighted according
to its overall contribution to the system. For each criteria a
score out of 10 is given for both solutions, where 10 indicates
less processing time, less labor, more coverage areas, less
operational cost and more environmental protection. Table I,
which has been developed in collaboration with Abu Dhabi
Municipality (ADM), shows the cost benefit analysis of using
drones compared to the existing methods. It’s clearly evident
from our informative estimates that the introduction of drones
will produce great benefits for the construction sector, even
with the most conservative estimates.
Our proposed technological solution, using the Site Inspec-
tion Drone (SID), provides a utilization of drone technology
in the construction industry. This paper is organized as follow:
Section II presents the typical work flow of the site inspection
and the proposed work flow via SID. Section III describes the
SID system architecture, component diagram, and the drone
hardware. Section IV describes the conducted experiment and
shows the efficiency of the results. Finally, Section V concludes
the paper, and highlights some future work to improve the
current system.
II. WORK FLOW OF T HE IN SP EC TI ON DRO NE
A. Current Inspection Scenario
According to the current ADM workflow, the inspection
is performed at different stages of the construction process
and at various time intervals. The current workflow, used
by the municipality for violation detection is displayed in
Fig. 1. Initially, all construction data and permits are stored in
the database of the Community Development Partners (CDP)
system. When the inspection schedule is due, the inspector
develops a site inspection plan based on the information stored
in the CDP system. Then, the inspector visits the site, collect
images and update the CDP system with the inspection details.
The inspector uses the collected images and data from the site
to detect violations. If a violation is detected, the inspector
issues warning/fines via the mobile application and update the
CDP system.
B. Proposed Inspection using Drones
Based on ADM regulations, three main inspection pro-
cedures have been identified: violation inspections, progress
of works inspections, and complaints assistance. The work
presented here will demonstrate only the first type of inspection
procedures, the violation inspections. Our new proposed drone
system will aid in detecting violations in the construction
process around the UAE.
A detailed description of how our system is used to check
violations is illustrated in Fig. 2 and Fig. 3. Initially, the drone
directly communicates with the municipality’s databases. The
system then creates schedules and timelines for the inspecting
drone. When the inspection time is due, the system sends an
2016 IEEE 59th International Midwest Symposium on Circuits and Systems (MWSCAS), 16-19 October 2016, Abu Dhabi, UAE
978-1-5090-0916-9/16/$31.00 ©2016 IEEE
TABLE I: Cost Benefit Analysis Developed with ADM
Criteria Criteria Weight Current Traditional Scenario Proposed Drone Scenario
Score (out of 10) Weighted Score Score (out of 10) Weighted Score
Processing Time 30% 4 1.2 8 2.4
Labors Required 10% 4 0.4 8 0.8
Coverage Area 20% 3 0.6 9 1.8
Operational & Maintenance Cost 20% 5 1 4 0.8
Environmental Protection 20% 5 1 8 1.6
Total 100% 21 4.2 37 7.4
Fig. 1: Current Inspection Scenario of ADM
Fig. 2: Drone scenario 1: Violation Inspection Diagram
Fig. 3: Drone Scenario 1: Violation Inspection Flow Chart
alert to an officer asking him to approve the inspection flight
using a drone. At that stage, the officer has to decide whether
the drone will perform the mission alone or be accompanied
by an inspector officer. If the officer approves the flight, the
drone will access the permits database to retrieve the GPS
Fig. 4: Component Diagram
coordinates of the construction site. After that, a flight route
will then be generated specific to that site, and the drone will
autonomously fly to the site, collect the required data, such
as live feed, panoramic images and in some cases 3D scans
of certain areas. The drone will then return to the take-off
position and will start recharging itself at the charging station
in preparation for another flight. Meanwhile, the drone will
upload all captured data to the database so that the officer at the
server station can remotely access the information. The officer
can then quickly review the data and decide if a violation
fine should be issued or not. The drone will be also sending
data periodically while on the mission, this will enable the
officer to view the live information from their smart mobiles
application or desktop computers. The inspection officer also
has the authority to send drones to a specific site at their own
will if needed, even if the scheduled time has not arrived.
III. SID SYS TE M ARC HI TE CT URE
A. Component Diagram
Fig. 4 illustrates the components diagram of our proposed
system. The system consists of three main components: the
municipality database, the server station and the drone. The
database information is sent wirelessly to the base station,
responsible for scheduling inspections, and defining the site
location and building codes that should be uploaded to the
drone before each inspection mission.
The drone system consists of four main blocks: the on-
board processor, on-board sensors, the autopilot system and the
2016 IEEE 59th International Midwest Symposium on Circuits and Systems (MWSCAS), 16-19 October 2016, Abu Dhabi, UAE
drone’s hardware. The on-board processor in the drones will re-
ceive the location information of the scheduled inspection visit.
Then, it will send this information to the autopilot system via
”Mavlink” [4] protocol to navigate to the desired destination.
Mavlink is a communication protocol developed specifically
for communicating with micro arterial vehicles. When the
drones reach the construction site, the on-board processor
will read the on-board sensors data, perform online analysis
on the images, store the results and sent them back to the
server station. The inspector at the server station then updates
the municipality database based on the requested mission. In
the current version of our drone, four on-board sensors are
used which are: RGB camera for capturing images/videos,
depth sensor for capturing 3D data of construction sites, laser
range finders for obstacle avoidance, and IR camera for night
inspection. All of the on-board sensors are USB connected
with the on-board processor.
B. System Architecture
Our proposed system architecture is shown in Fig. 5. The
monitoring ground station will communicate wirelessly with
both the municipality database and the drone’s on-board pro-
cessor. The monitoring ground station has server software that
includes: control panel, status panel, camera view and multi-
layer map. The drones is equipped with PIXHAWK autopilot
system that receives the desired drone location, reads and fuses
linear accelerations, angular velocities, earth magnetic fields
and barometric pressure to allow flight control. The autopilot
communicates with the on-board processor via ’Mavlink’. The
on-board processor is also connected to on-board sensors by
USB ports. The on board processor also includes a local path
planner that can assist in autonomous navigation, take-off and
landing.
Fig. 5: System Architecture
C. Drone Hardware
The frame of the drone is Flame Wheel ARF Kit [5]
where the propeller used in our drone is APC 10x4.7 Slo-
Flyer Pusher Propeller [6]. The drive motors that are used
to activate drone propellers are NTM Prop Drive motors [7]
and the battery system consists of a standard LiPo Battery
and a power regulator [8]. The drone uses ODROID-U3 for
processing sensor information and controlling output devices.
The processor is also connected to the Autopilot system via
MAVLINK protocol.
D. Ground Station Interface
In order to control the drone and monitor it’s activities, we
developed a web interfaces that runs on the on-board computer
and is served via an Apache web server [9] on the wireless
link as shown in Fig. 6. The web interface allows the inspector
to use any smart device (like a phone, tablet or computer) to
access the drone, control it’s mission, and monitor it’s progress
if needed.
Fig. 6: A Snapshot of our Developed Web Interface
E. Total Cost and System Limitations
The estimated cost of the proposed solution including
hardware cost, maintenance cost and operational cost is sum-
marized in Table II. Flight time duration is the main limitation
of our system. Due to our payload and power consumption (live
streaming of images, onboard image processing, ...), the flight
time of our drone is limited to 20 minutes. In addition, due to
the time constrain, inspecting multiple building would require
buildings being in a close proximity to each other.
IV. FLI GH T EXP ER IM EN T AN D PER FO RM AN CE ANALYS IS
A. Experiment Description
The live-demo conducted in ”Dubai Internet City”, an
outdoor area where the competition was being held, will be
used to demonstrate the capabilities of the proposed system.
A mockup construction site environment was developed by
the competition organization based on our request, and several
violations were embedded deliberately in order to test the SID
system’s ability to detect these violations live.
The SID starts by taking off from start position. It flies
autonomously to the inspection site. It then follows a prede-
fined set of GPS waypoints, uploaded to the drone as part of
the inspection mission. During the inspection process, images
are collected for offline 3D map reconstruction using Pix4D
2016 IEEE 59th International Midwest Symposium on Circuits and Systems (MWSCAS), 16-19 October 2016, Abu Dhabi, UAE
TABLE II: Total cost of the proposed solution
No. Item Cost (US)
1
Operational Cost:
- Charging stations cost
- Training cost
272
2
Maintenance Cost:
- Replacement of damaged parts
- Maintenance of mechanical parts
545
3
Hardware Cost: 3893
- Frame + Landing gears 353
- Propeller system 139
- Drive motors 82
- Power supply 19
- On-board processor 65
- Autopilot system + radio kit + GPS 476
- Ground Station open source
- RGB camera 293
- Depth Sensor 377
- Laser Finder 1491
- IR sensor 29
- Communication system 297
Other parts 272
Total Cost 4710
program and online violation detection. In addition, the system
onboard computer is connected with a database containing a
list of the violations to be detected. Moreover, using a web
interface, an observer in the ground station will monitor the
inspection process and will check and verify the highlighted
violation once detected. Finally, when the inspection process
is done, the drone flies back to its starting point and land. A
video of this work is available at [10].
B. Experiment Analysis
The position of the SID was recorded using the on-
board GPS unit, and the trajectory that the SID followed
during the inspection process from the takeoff position to the
landing position was shown in Fig. 7. Another example is
the online detected violation reported in the web interface as
demonstrated in Fig. 8. The officer can visualize the violation
via the web interface which as reported ”existing of worker
out of the working time”. Finally, using the collected images
during the inspection process, a 3D model of ”Dubai Internet
City” is generated offline using Pix4D program as shown in
Fig. 9. The green line shown in the 3D reconstructed images
shows the path of the Drone while taking images. The position
information was collected form the images where each image
has a GPS position stamp.
Fig. 7: Path of the Drone for Inspection Test
V. CON CL US IO N AN D FUT UR E WOR K
In this paper, the motivation, challenges and benefits of
the proposed SID were presented. A demonstration on how
Fig. 8: Detected Violation
Fig. 9: 3D Reconstructed Map of the Site being Inspected
the SID system can be integrated into the current workflow
of the ADM was presented. A comprehensive cost-benefit
analysis in collaboration with ADM that demonstrated how the
proposed SID system will present a cheap solution that vastly
increases efficiency, safety and practicality of the inspection
process was conducted. In future work, the flight time could
be significantly improved if high standard batteries are used. In
addition, although all the inspection scenarios proposed by our
system were performed during the day, our drone is equipped
with an infra-red camera that allows for the possibility of
performing the site inspection during the night time. Finally, a
list of the violations that should be detected could be provided
along with the qualitative and quantitative results based on
tests for each violation.
REF ER EN CE S
[1] L. Tung. Nokia puts telco drones to work inspecting
cell towers. [Online]. Available: http://www.zdnet.com/article/
nokia-puts- telco-drones- to-work- inspecting-cell- towers/
[2] R. Smith. Utilities turn to drones to inspect power lines
and pipelines. [Online]. Available: http://www.wsj.com/articles/
utilities-turn- to-drones- to-inspect- power-lines-and- pipelines-1430881491
[3] V. Khumalo. Boeing uses drones for inspection. [Online]. Available:
http://www.rocketmine.com/boeing-uses-drones-for-inspection/
[4] L. Meier. mavlink. [Online]. Available: http://wiki.ros.org/mavlink
[5] Flame wheel arf kit - features - dji. [Online]. Available: http:
//www.dji.com/product/flame-wheel-arf/feature
[6] Towerhobbies.com - apc 10x4.7 slow flyer pusher propeller. [On-
line]. Available: http://www3.towerhobbies.com/cgi-bin/WTI0001P?I=
LXBPPW&P=8
[7] Ntm prop drive series 28-30a 750kv-140w. [Online]. Avail-
able: http://www.hobbyking.com/hobbyking/store/ 16229 NTM
Prop Drive Series 28 30A 750kv 140w.html
[8] Dji officially released the landing gear for flame
wheel. [Online]. Available: https://www.dji.com/info/releases/
dji-officially-released-the-landing-gear-for- flame-wheel
[9] Apache http server project. [Online]. Available: https://httpd.apache.org/
[10] SIDTeam. Site inspection drone. [Online]. Available: https://www.
youtube.com/watch?v=jZS7cfHcJvY&feature=youtu.be
2016 IEEE 59th International Midwest Symposium on Circuits and Systems (MWSCAS), 16-19 October 2016, Abu Dhabi, UAE
... Remarkably, collaborative solutions (i.e., drones that operate simultaneously for a common goal, relying on exceptional communications between the different systems) have become increasingly popular in the last few years, with multimodal solutions showing promise as well. Drones have been typically employed in all sorts of inspection missions, including infrastructure, construction sites, radiology, industrial sites and stockpiles [73][74][75][76][77][78][79][80][81]. Within wind turbine inspections, there have been several technologies deployed, including edge analytics, digital twin technology, real-time anomaly detection using AI algorithms and autonomous drones [3,5,6]. ...
Article
Full-text available
With the ever-increasing demand for harvesting wind energy, the inspection of its associated infrastructures, particularly turbines, has become essential to ensure continued and sustainable operations. With these inspections being hazardous to human operators, time-consuming and expensive, the door was opened for drone solutions to offer a more effective alternative. However, drones also come with their own issues, such as communication, maintenance and the personnel needed to operate them. A multimodal approach to this problem thus has the potential to provide a combined solution where a single platform can perform all inspection operations required for wind turbine structures. This paper reviews the current approaches and technologies used in wind turbine inspections together with a multitude of multimodal designs that are surveyed to assess their potential for this application. Rotor-based designs demonstrate simpler and more efficient means to conduct such missions, whereas bio-inspired designs allow greater flexibility and more accurate locomotion. Whilst each of these design categories comes with different trade-offs, both should be considered for an effective hybrid design to create a more optimal system. Finally, the use of sensor fusion within techniques such as GPS and LiDAR SLAM enables high navigation performances while simultaneously utilising these sensors to conduct the inspection tasks.
... The data for these variables are collected using various techniques, including image processing, image registration, calibration, and correction [6]. In the study conducted by [7], a real-time robust system was created for inspecting construction sites, emphasizing the identification of violations. The system accomplished this goal by generating a 3D map of the construction site. ...
Conference Paper
Full-text available
This research introduces a Semi-Markov DecisionProcess (SMDP) approach to task allocation for UnmannedAerial Vehicles (UAVs) by considering the stochastic behaviorof battery levels. The SMDP allows the UAV to dynamicallyallocate tasks based on its current battery state and the timespent in each state. We compare the SMDP method with amanually assigned task sequence and demonstrate a significantreduction in completion time for a set of predefined tasks.The SMDP optimally assigns tasks, considering the stochasticnature of battery levels, resulting in improved efficiency andeliminating uncertainties associated with human allocation. Thefindings underscore the benefits of incorporating SMDP in UAVtask management, especially in scenarios dependent on real-time battery levels.
... It offers several advantages over other platforms, such as a simple structure, exceptional maneuverability, and relatively low maintenance cost. 1 Because of these advantages, multirotor UAVs have been adopted in various fields, including surveying, aerial photography, agriculture, mapping, and surveillance. [2][3][4][5] In addition to their current use, they hold great potential for further development in fields such as planetary exploration, delivery service, and construction. [6][7][8] As the utilization of UAVs continues to expand, the control will be increasingly demanded to be more sophisticated, and the operational environments will become more diverse. ...
Article
Full-text available
This study investigates the effects of rotor–rotor interaction on the wake and thrust characteristics of a small tandem rotor operating in a crosswind. Flow velocity and force measurements were conducted in a wind tunnel with two rotors arranged parallel to a crosswind. The results show that the rotor–rotor interaction significantly influences the wake characteristics and thrust generations of the tandem rotor and its effects vary depending on the crosswind speed and distance between rotor tips. In the tandem rotor configuration, the front rotor wake prevents the crosswind flow from reaching the rear rotor wake, thereby reducing the crosswind influence on it. However, under the strong rotor–rotor interaction, such as that caused by high crosswind speeds and short distances between rotor tips, the wakes of both rotors collide with each other and rapidly break down as they proceed downward. Tip-vortex characteristics are also affected by rotor–rotor interaction, which is investigated in terms of variations in the time-averaged tip-vortex trajectory and dissipation ratio with the strength of rotor–rotor interaction. These wake variations by rotor–rotor interaction lead to a decrease in thrust coefficients of the front and rear rotors, with a more significant reduction observed for the rear rotor. The thrust of the rear rotor is more significantly reduced as the crosswind speed increases and the distance between rotor tips narrows. This is mainly attributed to the increased axially induced velocity near the leading tips on the advancing side, retreating side, and centerline.
... Empirical studies on drone applications within the AEC industry have primarily focused on drone application areas, safety and health management, data processing and management, training aid, and flight control within the I4 framework [28][29][30][31][32][33][34][35][36][37], with only a few conducted within the I5 framework [11,16]. Sharma and Arya [11] used the IoT (Internet of Things) and a cloud server to implement a drone network to track landfill site air quality in real time in hilly areas or distant locations. ...
Article
Full-text available
Industry 5.0 is expected to solve the issue of job insecurity and reluctance to adopt emerging technologies in Industry 4.0 through increased workforce participation. To achieve that, proactive training should be afforded to enable the workforce to co-work with new technologies. Drones are an emerging technology increasingly adopted in construction, which has enormous data collection and safety implications when operational skills are inadequate. Yet, current drone training programs appear to be generic, and their ability to equip operators for flying tasks is questioned. This study aims to answer this question by assessing the adequacy of existing drone training programs and proposing training needs and strategies for drone operators in the construction industry. Data collected using semi-structured interviews from 22 purposively selected respondents in Hong Kong and desk research of websites were subjected to inductive content and thematic analysis using MAXQDA Analytics Pro 2022 software and narrative review. It was deduced that drone training courses in Hong Kong were mostly generic (UAS Pilot Training—Level 1, FPV operation). Interviewees considered existing drone training/training courses as inadequate for four reasons, including “lack of context-fitting considerations”, “incompetence of drone operators”, and “lack of demand/interest”. Regarding the need for specialized drone training courses, two barriers and two training ecosystem themes emerged, with the high cost of specialized courses being a potential barrier to adoption since they could hamper enrolment. The training ecosystem themes were consistent with the “the training is inadequate” theme. This study proposes two drone operation training strategies: “competence-based training” and a “train-the-trainer” model. Drone training courses or programs under each training strategy should include (i) training content such as “safety training” and off-GPS training and (ii) knowledge and skill maintenance measures such as mandatory continuous professional development and retraining techniques. The proposed training strategies will equip operators to work efficiently and safely with drones. The study offers valuable references for training organizations and government authorities.
Conference Paper
Full-text available
With the rise of renewable energy installations, in-line with the worldwide goals for sustainable energy, inspection became a crucial factor for the longevity of these systems. This is especially true for wind turbines where the system performance is impacted by the condition of the system's components, particularly the blades. However, human inspections are dangerous and costly, suggesting that robots can enhance such operations. That said, the need for multiple robots to undertake several types of inspection creates other issues such as drone swarm communications and additional costs. As such, this work aims to introduce a multi-modal design for a robot capable of inspecting multiple areas on the wind turbine independently. The design was tested using Solidworks simulations for stress analysis, demonstrating sufficient robustness and maintaining its overall form under different robotic configurations. The paper also highlights the aerodynamic modelling of the robot using Ansys Fluent at high-speed conditions expected in wind farms. Finally, we highlight the different mechanisms and locomotion methods required for convenient operation of the robot while testing them within a simulated environment, illustrating their feasibility in the required environment. I. Nomenclature = Identity matrix = Turbulence kinetic energy (J/kg) = Characteristic length of the airfoil (m) p = Pressure (Pa) = Reynolds number S = Source term (kg/m 3) = Inlet velocity of the fluid (−1) U = Flow velocity (m/s) + = Distance from the first grid cell to the surface = rate of dissipation of turbulent kinetic energy (2 −3) = Fluid viscosity (Pa s) = Fluid density (kg/m 3) = Stress tensor (N/m 2) = specific turbulence dissipation rate (−1)
Chapter
The rapid evolution of drone technology has revolutionized data acquisition in the construction industry, offering a cost-effective and efficient method to monitor and map engineering structures. However, a significant challenge remains in transforming the drone-collected data into semantically meaningful 3D models. 3D reconstruction techniques usually lead to raw point clouds that are typically unstructured and lack the semantic and geometric information of objects needed for civil engineering tools. Our solution applies semantic segmentation algorithms to the data produced by NeRF (Neural Radiance Fields), effectively transforming drone-captured 3D volumetric representations into semantically rich 3D models. This approach offers a cost-effective and automated way to digitalize physical objects of construction sites into semantically annotated digital counterparts facilitating the development of digital twins or XR applications in the construction sector.
Chapter
Housing inspection is an essential task to ensure quality delivery during construction operations, in accordance with relevant building manuals and standards. It is critical to ensure that customers are satisfied with housing delivery by monitoring construction activities per stage inspections. The evidence gathered from housing construction inspectors as well as from previous studies suggests that the effectiveness and efficiency of inspection processes currently in the South African construction industry are unsatisfactory. This paper reports on South African research that explores the feasibility of adopting an automated inspection approach to incorporate digital technologies such as 3D laser scanners, drone technologies, and Building Information Modeling (BIM) into the housing construction inspection process to facilitate more effective and efficient data collection, processing, and reporting for improvement of the inspection process. The paper also discusses the interview results of housing construction inspectors about the possibility of adopting an automated inspection approach. The results show that the automated approach with the new digital technologies has the potential to improve housing construction inspection for the delivery of quality housing. The research efforts will also enhance the current knowledge of the housing construction inspection process.
Utilities turn to drones to inspect power lines and pipelines
  • R Smith
R. Smith. Utilities turn to drones to inspect power lines and pipelines. [Online]. Available: http://www.wsj.com/articles/ utilities-turn-to-drones-to-inspect-power-lines-and-pipelines-1430881491
Boeing uses drones for inspection
  • V Khumalo
V. Khumalo. Boeing uses drones for inspection. [Online]. Available: http://www.rocketmine.com/boeing-uses-drones-for-inspection/
Nokia puts telco drones to work inspecting cell towers
  • L Tung
L. Tung. Nokia puts telco drones to work inspecting cell towers. [Online]. Available: http://www.zdnet.com/article/ nokia-puts-telco-drones-to-work-inspecting-cell-towers/
Available: http://wiki.ros.org/mavlink [5] Flame wheel arf kit-features-dji
  • L Meier
  • Mavlink
L. Meier. mavlink. [Online]. Available: http://wiki.ros.org/mavlink [5] Flame wheel arf kit-features-dji. [Online]. Available: http: //www.dji.com/product/flame-wheel-arf/feature
com -apc 10x4.7 slow flyer pusher propeller
  • Towerhobbies
Towerhobbies.com -apc 10x4.7 slow flyer pusher propeller. [Online]. Available: http://www3.towerhobbies.com/cgi-bin/WTI0001P?I= LXBPPW&P=8
Site inspection drone
  • Sidteam
SIDTeam. Site inspection drone. [Online]. Available: https://www. youtube.com/watch?v=jZS7cfHcJvY&feature=youtu.be