Available via license: CC BY 4.0
Content may be subject to copyright.
International Journal of Business and Management; Vol. 19, No. 6; 2024
ISSN 1833-3850 E-ISSN 1833-8119
Published by Canadian Center of Science and Education
49
Enhancing International Trade Security: Real-Time Risk Assessment
in Brazilian Customs with Blockchain Technology
Daiane Rodrigues dos Santos1
1 Rio de Janeiro State University, Rio de Janeiro, Brazil
Correspondence: Daiane Rodrigues dos Santos, Rio de Janeiro State University, Rio de Janeiro, Brazil. E-mail:
daiane.santos@uerj.br
Received: August 3, 2024 Accepted: September 2, 2024 Online Published: September 20, 2024
doi:10.5539/ijbm.v19n6p49 URL: https://doi.org/10.5539/ijbm.v19n6p49
Abstract
This study introduces an innovative approach by integrating blockchain technology with risk analysis methods
for international trade. The practical application and tangible improvements observed in a real case study
highlight the originality and value of this approach, offering a robust framework for future research and practical
implementation. The security of exports presents a challenge for customs authorities, requiring the analysis of
vast volumes of data from diverse sources and stakeholders. Blockchain technology offers a viable solution by
facilitating the sharing of reliable and secure data among participants in the international supply chain. This
study proposes a distributed risk analysis method based on Blockchain to detect and prevent fraud and
irregularities in Brazilian exports. The technique aims to enhance efficiency and protection in customs clearance,
providing real-time risk analysis feedback at various supply chain stages. The method includes four stages: data
collection, risk classification, validation, and feedback. It employs a Model Form for Insertion into the
Blockchain, the DBSCAN algorithm to identify out-of-pattern transactions (outliers) at the manufacturing stage,
and a Monte Carlo simulation to analyze the volume and specific products exported from Rio Grande do Sul. We
applied this method in a case study of beef exports from Brazil. Using this method led to significant
improvements in processing times, operational costs, transparency, and security in the supply chain. The
blockchain-based approach identified fraudulent activities, such as value tampering and undeclared goods,
enhancing compliance and trust among trade partners. Implementing this method can significantly reduce
operational costs and processing times in international trade while increasing security and transparency. This
approach benefits customs authorities and private operators, promoting a more reliable and efficient commercial
environment. Adopting blockchain technology in international trade can reduce fraudulent and illicit practices,
promoting greater fairness and equity in commercial transactions. Increased transparency strengthens trust
among various actors in the supply chain, fostering a sense of social responsibility.
Keywords: blockchain, export, security, customs, risk analysis
1. Introduction
The Fourth Industrial Revolution, spotlighted at the 2016 World Economic Forum, has underscored the profound
impact of technologies such as Artificial Intelligence (AI), blockchain, and big data across the globe. To remain
competitive in an increasingly demanding, connected, and dynamic global market, businesses and professionals
must embrace these disruptive technologies. As automation and AI take over repetitive and predictable tasks, the
human workforce must evolve, fostering creativity, complex problem-solving abilities, emotional intelligence,
and critical thinking as core competencies. In this context, adaptability becomes a crucial competitive edge.
Leveraging technology goes beyond merely enhancing production and operational efficiency; it involves seizing
new business opportunities and innovative service models that emerge at the intersection of various
technological domains. Preparing for this new era demands technical proficiency and strategic foresight,
requiring a mindset geared towards continuous learning, reinvention, experimentation, and interdisciplinary
collaboration on a global scale.
International trade, a cornerstone of the global economy, generated over 25 trillion dollars in 2023, providing
employment and income for millions. However, exporting and importing goods presents numerous challenges
and risks for private operators and public authorities. These risks include fraud, irregularities, sanitary and
environmental regulations violations, smuggling, terrorism, and organized crime. Such factors compromise the
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
50
security, efficiency, and competitiveness of international trade, underscoring the urgent need for coordinated and
integrated actions among various stakeholders in the supply chain.
Customs agencies play a pivotal role in the international trade landscape, controlling and supervising the entry
and exit of goods in countries (Howse & Trebilcock, 2005). They ensure compliance with national and
international laws and regulations, collect taxes, and prevent and suppress illicit activities (Tacconi & Brack,
2003). To fulfill these responsibilities, customs officials analyze vast volumes of data from diverse sources,
including documents, declarations, computerized systems, physical inspections, scanner images, and cameras.
Additionally, customs agencies communicate and cooperate with other public bodies, such as regulatory agencies,
police forces, intelligence services, and private operators, including exporters, importers, carriers, customs
agents, and banks.
Efficiency and innovation hold paramount importance in the contemporary landscape of customs operations.
Stakeholders dealing with a vast array of information and numerous participants require robust support to
manage their activities effectively (Huria, 2019). The rapid and secure data processing and analysis are crucial
for surveillance, tax collection, and combating illegal activities. Within this framework, blockchain technology
emerges as a transformative tool, significantly enhancing the efficiency and security of international trade. It
facilitates the exchange of reliable and secure data among supply chain participants, creating distributed,
verifiable, and immutable records that multiple authorized parties can update and access without needing a
central intermediary.
Blockchain technology's role in enhancing efficiency provides a promising outlook for the future of customs
operations. It facilitates the tracking and transparency of trade operations, reduces costs and processing times,
increases security and trust among agents, and introduces new methods for payments and financing. Additionally,
blockchain aids in export risk analysis by identifying and preventing potential fraud and irregularities, such as
value manipulation and undeclared products. In the corporate context, technology plays a crucial role in financial
procedures and contract management, which have evolved into the "smart contract" model, as described by Lima,
Hitomi, and Oliveira (2018, p. 6). These contracts, known for their efficiency, are part of a broader distributed
ledger system, with transparency and immutability ensured by blockchain technology. Calixto (2019) highlights
that blockchain extends further, possessing the capability to securely share and transfer digital assets, thereby
eliminating risks of duplication or fraud in transactions. This efficient 'smart contract' model instills confidence
in the future of financial procedures and contract management.
This article proposes a distributed risk analysis method for customs based on blockchain, aiming to detect and
prevent export fraud and irregularities. The method involves four stages: data collection, risk classification, data
validation, and feedback. Data collection entails recording relevant information about export operations on the
blockchain through smart contracts, which are self-executing codes defining the transactions' rules and
conditions. Risk classification uses machine learning algorithms to assign risk levels to operations based on
pre-established criteria, such as the operators' history, the origin and destination of products, and the value and
weight of the goods, among others. Data validation involves verifying the authenticity and consistency of the
data recorded on the blockchain through consensus mechanisms and protocols, ensuring all network participants
have the same data version. Feedback provides information on the risk analysis outcome to involved parties,
such as customs, exporters, and importers, allowing for appropriate decision-making and actions.
The proposed method was applied to a case of beef exports from Brazil, demonstrating its potential effectiveness
and benefits. Beef is a high-value product with significant commercial interest but is also subject to sanitary and
phytosanitary risks, which can affect public health and the environment (Abdis et al., 2023). Applying the
method to Brazilian beef exports illustrates how technology can adapt to meet specific sector challenges.
Additionally, the method presented in this article can be adapted and applied to other products and trade routes,
improving efficiency and protection in international trade. The potential of blockchain can also be used for other
products and across various products simultaneously, thereby highlighting its capacity to significantly optimize
global supply chains and strengthen international trade relations.
This article contributes to the field of customs risk assessment by introducing a novel methodology that
integrates advanced clustering algorithms with blockchain technology, specifically tailored for the complex
landscape of international trade. By leveraging the DBSCAN algorithm, this approach effectively identifies
outlier transactions within customs data, addressing the critical need for robust anomaly detection in trade
operations. The methodology's application to Brazilian beef exports exemplifies its practicality. It demonstrates
how it can enhance decision-making processes by providing a data-driven risk score incorporating historical
export behaviors and market dynamics. This integration strengthens the transparency and traceability of
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
51
transactions through blockchain and offers a scalable solution adaptable to various commodities and trade
environments. Theoretical foundations support this method through established principles of risk assessment,
utilizing statistical measures such as deviation from the mean and standard deviation, which are widely
recognized in finance and supply chain management. This approach empowers customs authorities to manage
and mitigate risks proactively, ensuring compliance and safeguarding economic interests. By bridging the gap
between theoretical models and real-world applications, this study provides a comprehensive framework that
advances the efficacy and reliability of customs risk management systems.
Additionally, this article presents a method for integrating customs information into blockchain systems through
a simplified pseudocode approach. By detailing a structured model for data insertion, the paper provides a
practical framework that enhances the transparency and security of customs transactions. The pseudocode
outlines a step-by-step process for encoding critical shipment details—such as order identification, manufacturer
location, product specifications, and shipping terms—into blockchain entries. Using blockchain's immutable
ledger, the proposed system protects data integrity and streamlines the risk assessment by enabling efficient
tracking and analysis of goods throughout the supply chain. This innovative approach bridges the gap between
theoretical blockchain applications and practical implementation in customs operations, offering a robust
solution to modernize and fortify international trade practices.
Moreover, applying this method to Brazilian beef exports demonstrates its practical utility and adaptability to
sector-specific challenges. The study highlights the potential of blockchain technology to optimize global supply
chains and strengthen international trade relations, offering a robust solution to modernize and fortify trade
practices. By integrating theoretical and practical insights, this research provides a comprehensive framework
that advances the efficacy and reliability of customs risk management systems, positioning customs
administrations at the forefront of innovation in the digital era.
This article, comprising eight sections, addresses various aspects of customs risk analysis using blockchain
technology. In "Blockchain Technology," section two, we explore the characteristics and advantages of
blockchain, emphasizing its applicability in managing customs risks. The "Proposed Method for Customs Risk
Assessment" section details the developed methodology, including data collection, risk classification, data
validation, and feedback. Following this, the "Manufacturer Risk Assessment" and "Logistics Agent Risk
Assessment," sections 4 and 5, present specific case studies to illustrate the practical application of the method.
The "Classification and Integration with Blockchain," section six, discusses the methodology implementation
within the blockchain environment, highlighting the security and efficiency benefits. Finally, the "Final
Considerations" section synthesizes the study's main conclusions. It suggests future directions for research and
practical applications, instilling confidence in the proposed methodology's potential.
2. Blockchain Technology
One potential application of Blockchain technology is in managing customs risks, aiming to ensure the security
and compliance of foreign trade operations. Customs risks can range from non-compliance with tax, tariff, and
sanitary regulations to practices like smuggling, counterfeiting, and terrorism (Nikoofal et al., 2023). Customs
risk analysis involves identifying and assessing potential import and export operations risks and defining
appropriate control and inspection measures to mitigate them.
According to Haque et al. (2024), blockchain technology is a decentralized database that operates without a
central authority or third-party verification. It consists of a series of interconnected blocks containing a hash of
the previous one, forming a continuous chain from the initial or "genesis" block to the most recent one.
Blockchain technology monitors, controls, and secures Internet of Things (IoT) devices, using decentralization
and encryption for protection. Blockchain transactions do not require intermediaries, making them trustworthy,
and the technology offers features like decentralization, immutability, and transparency, which provide
significant benefits in terms of increased security and data protection.
Blockchain also serves as a verification system that, among other things, records the sale/purchase (ownership)
of cryptocurrencies like Bitcoin, Ethereum, and Ripple. For this reason, distributed ledgers in the form of
Blockchain promise to be a transformative technology that will revolutionize the business world as we know it.
This technology allows for cryptocurrency ownership records and running different types of applications,
platforms, information storage, and distribution systems (Haile, 2024). At the governmental level, Blockchain
can monitor and control specific tasks, such as voting systems, tax collection, passport issuance, real estate
registration, delivery of grants, and other benefits (Santos, 2023). Blockchain is a distributed network of multiple
computers in different locations. They can exchange, receive, and store value or information with each other.
Each transaction is shared across the entire network and is recorded in a "block" when the whole network
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
52
confirms that the transaction (information) is valid, using the transactions (past information) from previous
blocks.
Permissioned Blockchain is a technology where only authorized agents can participate. According to Rizzardi et
al. (2022), permissioned Blockchain is the technology that allows only authorized agents to join, creating a
network environment where an entity or consortium strictly controls access and permissions. The restrictive
nature of permissioned Blockchains brings privacy and operational efficiency benefits, making them particularly
interesting for business and governmental organizations that require high levels of security and control over their
transactions and data. The public Blockchain operates as a decentralized network comprising Blockchain storage
entities. Each of these entities has a comprehensive replica of the entire system. This approach ensures the
system's resilience if many network nodes become inaccessible and data is lost. The entire system can be
reconstructed using a single node that maintains a complete copy of the Blockchain (Haque et al., 2024).
Blockchain technology-based networks generally fall into two main types: private (permissioned) and public
(non-permissioned). However, beyond these two types, there is also a third type, called hybrid or consortium
blockchain, which mixes features of both private and public blockchains. In this variant, access is controlled, and
the network's distributed access is restricted to a limited number of participants. It is also worth noting that the
governance of a consortium blockchain is shared among the founding members, allowing for a more democratic
and efficient network management. This contrasts with public blockchains, where governance is decentralized,
and with private blockchains, where governance is centralized in the entity operating the network.
Blockchain technology offers an innovative approach to managing customs risks, providing a secure, transparent,
decentralized platform for monitoring and controlling foreign trade operations. Blockchain operates without
intermediaries and, combined with its features of immutability and distributed verification, ensures greater
security and reliability in transactions. Additionally, the distinction between permission and public blockchains
and introducing hybrid or consortium blockchains allows for application flexibility. This flexibility can meet the
security and privacy needs of business and governmental organizations while also addressing the resilience and
transparency required by decentralized networks.
3. Enhancing Customs Risk Management through Blockchain Integration
Customs risk assessment, a critical element in safeguarding the integrity of international trade operations, is on
the brink of a transformation. Traditionally, customs risk assessment methods rely on statistical analyses and
predictive models that utilize historical data to identify risk patterns, Berk e Bleich (2013). While effective in
certain aspects, these methods face significant challenges, such as managing large volumes of real-time data and
integrating information from multiple sources. However, the future holds promise with the potential of more
transparent and efficient customs risk assessment techniques, Okazaki, (2017). These techniques could not only
overcome the current challenges but also foster trust and collaboration between trade partners and customs
authorities.
Recent literature underscores the transformative potential of emerging technologies like blockchain in customs
risk assessment. Gao et al. (2018) highlight blockchain's ability to address the challenges of traditional methods
by providing a decentralized and secure platform for information sharing. Blockchain technology enables the
creation of immutable and auditable records accessible in real time by all stakeholders. This plays a crucial role
in enhancing transparency and trust, thereby reassuring the audience about the benefits of blockchain in customs
risk assessment. It also facilitates the early detection of risks such as fraud and discrepancies in customs
documentation. Therefore, applying blockchain in customs risk assessment represents not merely a technological
innovation but a necessary evolution to tackle the challenges of modern global trade.
In the Results section, summarize the collected data and the analysis performed on those data relevant to the
discourse that is to follow. Report the data in sufficient detail to justify your conclusions. Mention all applicable
results, including those that run counter to expectation; be sure to include small effect sizes (or statistically
nonsignificant findings) when theory predicts large (or statistically significant) ones. Do not hide uncomfortable
results by omission. Do not include individual scores or raw data, except for single-case designs or illustrative
examples. In the spirit of data sharing (encouraged by APA and other professional associations and sometimes
required by funding agencies), raw data, including study characteristics and individual effect sizes used in a
meta-analysis, can be available on supplemental online archives.
Customs risk, a term encompassing potential threats and vulnerabilities in international trade and customs
processes, has far-reaching implications. These risks, including illicit practices and procedural deviations, can
compromise the integrity of cross-border commercial operations, posing threats to national security, public
health, and economic stability. For instance, manipulating goods' value in customs documents, a common
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
53
strategy to evade taxes, generates significant government revenue losses. Similarly, the smuggling of prohibited
goods without declaration poses imminent risks to national security, public health, and the economy, given their
ability to illegally cross borders (Juma et al., 2019).
Customs administrations are essential to reducing the risk associated with customs, Karklina-Admine et al.
(2024). However, the effectiveness of international trade is harmed by obstacles they must overcome, such as
documented fraud and non-compliance with customs regulations, which can lead to fines, delays in customs
clearance, and supply chain disruptions. Import or export of goods that pose security risks, such as weapons or
hazardous materials, also constitute significant dangers to public safety and national protection. In response,
customs administrations implement risk assessment methods and advanced technologies to detect and mitigate
such risks, ensuring compliance with trade regulations, protecting national interests, and facilitating legitimate
trade activities.
The proposed method's customs risk assessment consists of two main steps: transaction representation and
subsequent classification. This process is executed simultaneously in two different phases of the supply chain:
first, in the goods' manufacturing phase and then in the transportation phase.
The procedure is initially activated in the production phase, at which point the risk assessment process begins
immediately after the importer uploads the order details into the Blockchain system. The uploaded information
includes, but is not limited to, the manufacturer's location, invoice details, and the product's country of origin.
This step primarily aims to identify potential risks of value manipulation, such as under- or over-invoicing,
through careful analysis of the submitted transactions, which will be discussed in the next section.
Table 1. Model Form for Insertion into the Blockchain (Phase 1)
Information Field Detailed Description
Order ID A unique identifier for each order, facilitating traceability and cross-referencing.
Order Date The date the order was placed, including day, month, and year.
Manufacturer Location The complete address of the manufacturing location, including city, country, and, if
applicable, industrial zone.
Invoice Details Invoice number, date, detailed description of items, quantities, and unit prices.
Country of Origin The country where the products were manufactured or produced.
Product Description A complete description of the products, including categories, technical specifications,
and HS codes.
Declared Value The total declared value of the products, as stated on the invoice.
Shipping Terms Shipping conditions, including, for example, FOB, CIF, and carrier information.
Estimated Date of
Arrival The expected date for the products to arrive at the final destination.
Associated
Documentation
References to associated documents, such as certificates of origin, export/import
licenses.
Source. Own Elaboration.
Once completed, this model form (Table 1) would be encoded in JSON (JavaScript et al.), for example, and
inserted into the Blockchain chosen by customs as a transaction. Each field in Table 1 serves as a crucial piece of
information that contributes to the transparency, security, and efficiency of the import/export process, allowing
all stakeholders to verify the authenticity and compliance of the information without compromising data
confidentiality.
Moving forward in the supply chain, the second phase is the transportation phase of the goods, where the
transport agent begins their risk assessment as soon as the relevant information becomes available on the
Blockchain. It is important to note that Blockchain technology can significantly benefit the supply chain in
various ways: Enhanced transparency, improved traceability, efficiency, intelligent contracts, security, cost
reduction, and real-time tracking (Casado-Vara et al., 2018). According to the authors, Blockchain technology
can potentially revolutionize the supply chain industry by improving transparency, traceability, security,
efficiency, and cost-effectiveness (Deshpande, 2012).
According to Dujak and Sajter (2019), Blockchain technology promises to overcome trust issues by enabling a
trustless, secure, and authenticated system for exchanging logistics and supply chain information in supply
networks. New implementations in the supply chain are evolving from Blockchain to a broader notion of
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
54
distributed ledger technologies. Blockchain introduces an innovative trust mechanism based on advanced
cryptographic technology and a decentralized network, eliminating the need for trust in intermediaries. This
revolutionary feature of Blockchain radically transforms relationships within the supply chain, which
significantly depend on mutual trust between partners. Furthermore, the blockchain network facilitates the secure
and transparent sharing of information among all supply chain partners, enhancing relationship improvement and
operational efficiency (Wang et al., 2020).
The second phase uses data such as the shipping route and specific details of the commercial transaction to
identify risks associated with undeclared goods. Table 2 presents a model for inserting information related to the
supply chain part into the Blockchain.
Table 2. Model Form for Insertion into the Blockchain (Phase 2)
Information Field Detailed Description
Transportation ID A unique identifier for the transportation operation, facilitating traceability and
cross-referencing.
Order ID A unique identifier is associated with the original order, linking the transportation
operation to the specific order.
Start Date of Transportation The date when the transportation of the goods begins is formatted as
YYYY-MM-DD.
Transportation Agent Information about the logistics company responsible for the transportation,
including name, contact, and address.
Shipping Route Details about the origin and destination of the goods, including cities, countries,
and specific ports or airports.
Estimated Time Estimated time for the delivery of the goods, expressed in days, weeks, etc.
Transaction Details Information about the shipping terms (e.g., CIF, FOB), declared value, and
insurance details.
Packing List A detailed list of all transported items, including identification, description,
quantity, weight, and dimensions.
Transportation
Documentation
References to key documents for transportation, such as the Bill of Lading and the
Packing List.
Safety Observations Notes on safety requirements, hazard classifications of the products, and any
required inspections.
Company Route Frequency
Information on the volume of contracts of the company responsible for the
movement of the goods. Example, whether it is an entering or continuing
company. Regularity information can be added to the form.
Insurance Details about the insurance policy covering the goods, including insured value,
insurance company, and policy number.
Source. Own Elaboration.
The analysis conducted at each of these phases is carried out on two distinct levels: an individual and a global
level. At the individual level, the method examines the shipment under review, using the historical transactions
of the same importer for comparison purposes. At the global level, the total set of available transactions,
covering all importers, is used as a basis for comparison. The evaluation combines the results obtained at both
levels in each phase to determine if the current shipment presents risks. This process involves modeling the
history of shipments as points in a multidimensional space, with each phase of the evaluation divided into risk
and safe subspaces. The risk subspace encompasses the shipments where risk is confirmed, while the safe
subspace includes those without identified issues.
To illustrate the practical application of the risk assessment process in the logistics industry, let's consider a
real-world example. Imagine a logistics company that uses a data-based risk assessment system to monitor and
manage international shipments. This system collects data from various sources, including information about the
sender, the recipient, the nature of the goods, the history of previous shipments, transport routes, and transit
times. This data is then used in the evaluation phase (Data Collection and Risk Analysis), which is followed by
the evaluation process and the decision-making based on the assessment. Additional verification measures are
recommended if the goods fall into Group A - a group with initially assessed risk. This could include a more
detailed inspection of the goods and documentation or the consideration of additional insurance or alternative
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
55
transport routes. On the other hand, if the goods fall into Group B – a group without initially assessed risk, the
shipping process proceeds without the need for extra interventions, as it is classified as safe based on historical
analyses.
At each evaluation step, the results of the two levels of analysis are represented as sets of discrepant factors,
individual and global. These factors are then processed to determine the current shipment's risk classification,
considering the specific perspective of the evaluation phase in question. Besides the safe and risky classifications,
a shipment can also be categorized as indeterminate. Shipments with this classification require a reevaluation
through a centralized risk assessment process conducted by the customs administration. The introduction of the
classification of shipments as indeterminate aims to enhance the accuracy of the proposed classification method.
For this, a threshold is introduced that establishes the minimum difference required between the discrepant
factors for the classification of the shipments. This threshold is of utmost importance, as shipments with a small
gap between outliers do not present distinct characteristics enough for a direct classification, requiring further
investigation by the customs authorities.
Figure 1 illustrates the phases of the life cycle of a customs risk assessment. Each phase involves different
activities and decisions. This study proposes applying a risk assessment methodology in phases 1 (manufacturing)
and 2 (transporting goods) of the export process. After classifying the risk level A or B of the shipment on the
Blockchain, a message would be sent to the customs authorities. The customs authorities, as key stakeholders in
the risk assessment process, would then act according to the risk level of the goods. This highlights the practical
implications of the proposed methodology and its potential to enhance the accuracy of the classification method.
Figure 1 - Customs Risk Assessment: Proposed Process
Source. Own Elaboration.
The proposed method for customs risk assessment, rooted in blockchain technology, represents a groundbreaking
innovation in risk management for international trade in the digital era. The integration of blockchain enhances
data transparency and security and fosters a collaborative environment among various economic agents in the
supply chain, including manufacturers, logistics agents, and customs authorities. This method significantly
improves operational efficiency and accuracy in risk identification and strengthens the integrity and reliability of
cross-border commercial operations. Ultimately, adopting this technological approach positions customs
administrations at the forefront of innovation, enabling them to confront the complex challenges of
contemporary global trade with greater resilience and effectiveness.
Integrating blockchain into the customs risk assessment process significantly advances over traditional methods.
By enhancing data transparency, improving security, and fostering collaboration among supply chain agents, the
proposed method directly addresses the limitations of current systems. The method's validity and reliability gain
reinforcement from blockchain’s ability to provide immutable and auditable records, which are crucial for trust
and efficiency in customs operations, Rouhani e Deters (2021). As international trade continues to evolve,
innovative methods like this are essential to ensure customs administrations can effectively respond to the
increasing complexities of the global trade environment. Embracing this technological approach positions
customs administrations at the forefront of innovation, enabling them to confront contemporary challenges with
greater resilience and efficacy.
4. Manufacturer Risk Analysis in the Export Industry
The proposed method relies on a robust dataset on the export of beef (including offal, preparations, and preserves)
from various cities in the State of Rio Grande do Sul, Brazil, for 2018 (Note 1). This experiment utilized data
from 23 large-scale companies that exported an average of USD 3.74 per kilogram of beef. The data collected
from Funcexdata in April 2024 is a reliable source. The companies are located in nineteen cities: Santo Ângelo,
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
56
Santa Rosa, Porto Alegre, São Luiz Gonzaga, Marau, Rio Grande, São Sebastião do Caí, Pelotas, Encantado,
Caxias do Sul, Passo Fundo, Garibaldi, Montenegro, Trindade do Sul, Estância Velha, São Gabriel, Bagé,
Alegrete, Estação, and Frederico Westphalen.
Graph 1 illustrates the volume and average price data for beef (including offal, preparations, and preserves)
exported from Rio Grande do Sul from 2010 to March 2024 (HS 0201). This comprehensive analysis, which
includes significant fluctuations in both value (USD FOB) and average price (USD FOB/Kg), provides a detailed
understanding of the beef trade evolution. The data shows a trajectory marked by significant fluctuations in value
(USD FOB) and average price (USD FOB/Kg). Starting in 2010 with an export value of USD 198,817,790 and
an average price of USD 3.36 per kilogram, a considerable increase was noted in 2011, reaching an export value
of USD 247,254,609 and an average price of USD 4.42 per kilogram. After this peak, the value and average
price experienced variations, with periods of decline and recovery reflecting global market fluctuations and
internal production conditions. From 2016 onwards, a trend of recovery and continuous growth is observed until
2022, with a notable peak in 2022, where the export value reached USD 442,962,338 and the average price hit
USD 5.34 per kilogram, the highest recorded in the analyzed period. However, in 2023, a significant reduction in
export value and average price is observed, indicating possible challenges in the international market or changes
in production and export conditions.
Figure 2. Beef Export (including offal, preparations, and preserves) from Rio Grande do Sul, Brazil, between
2010 and March 2024
Source: Own elaboration based on Funcex data.
To evaluate the risk of changes in import values, previous shipments are placed in a three-dimensional space,
where each shipment represents a point with cost and manufacturer location coordinates. All companies exported
beef (including offal, preparations, and preserves) in 2018. Graph 2 illustrates the dispersion of these shipments.
The export trade flow data for beef (including offal, preparations, and preserves) from the Rio Grande do Sul in
2018 appears in Table 3, which details the contribution of companies of various sizes to the total export value
(USD FOB) and the number of participating companies. The microenterprise segment contributes a small share,
with an export value of USD 84,598, involving only two companies. Small companies contribute a larger share,
with 26 companies exporting USD 328,731. Medium-sized companies also make a significant contribution, with
five companies exporting a total of USD 2,853,180.
0
1
2
3
4
5
6
0
50,000,000
100,000,000
150,000,000
200,000,000
250,000,000
300,000,000
350,000,000
400,000,000
450,000,000
500,000,000
Value (US$ FOB) Average price (US$ FOB/Kg) - secondary axis
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
57
Table 3. Export from Rio Grande do Sul – Brazil of Beef (including offal, preparations, and preserves) – Most
recent public data
DESCRIPTIONS 2018
Companies classified by
size or firm size Value (USD FOB)
Micro company 84,598
Small 328,731
Special 8,122,678
Medium 2,853,180
Large 227,577,594
Not Classified 235
.
Figure 3. Previous Shipments in Three-Dimensional Space
Source. Own elaboration.
The risk analysis strategically uses these coordinates to check for inconsistencies or changes in declared values
that might suggest an intention to reduce the amount of taxes owed. By analyzing the relationship between the
cost of the goods, the location of the manufacturer, and the code referring to beef (including offal, preparations,
and preserves), it is possible to verify if there have been changes in the information provided by the importer,
which could compromise the integrity and legality of the commercial transaction.
4.1 Identification of Outlier Transactions at the Manufacturer Stage
It is important to note that a deviation can be, for example, the correct payment of taxes. Customs duties are an
important source of revenue for the country, and sellers who do not pay them cause several problems. When
sellers do not pay customs duties, the government loses a significant source of revenue for the country. Sellers
who evade customs duties can offer their products at a lower price than their competitors who follow customs
regulations. This harms companies that comply with the rules and creates unfair competition. When sellers do
not pay customs duties, goods take longer to be cleared, which disrupts trade. Sellers not paying customs duties
may encourage illegal trade, further affecting the country's economy. In summary, sellers' non-payment of
customs duties significantly impacts the country's economy and trade (Dangsawang & Nuchitprasitchai, 2024).
Graph 3 presents all previous shipments in three-dimensional space with an outlier (Average price of USD 2.5
FOB/Kg). One of the objectives of this article is to present a methodology capable of identifying potential
outliers (information that deviates from the rest) and serving as supporting information for customs to carefully
investigate this information recorded on the Blockchain via form number 1 (Declared Export Value).
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
58
Figure 4. Previous Shipments in Three-Dimensional Space with Outlier (lower cost than the others)
Source. Own elaboration.
The figure 4 illustrates the cost distribution associated with beef shipments from various locations near Porto
Alegre, represented in a three-dimensional graph that combines geographic coordinates (longitude and latitude)
with the declared costs of each shipment. Among the colored points, which indicate regular shipments with cost
variations within an expected range, one point stands out as an outlier. This point symbolizes a shipment with a
declared cost significantly lower than the other points (average price of $2.5 FOB/Kg), marking it as a clear
outlier in cost distribution. Identifying this point is crucial for customs and risk analysis, as it suggests a possible
discrepancy or irregularity, such as under-invoicing or fraud in value declaration. Compared to what other
shipments regularly declared, the much lower cost highlights the importance of a detailed investigation into this
specific shipment to understand the reasons behind the discrepancy in declared costs and take the necessary
corrective measures to ensure compliance and protect economic and regulatory interests.
To detect points like the one presented in Graph 3, this article proposes an algorithm based on point distance,
DBSCAN (Density-Based Spatial Clustering of Applications with Noise). This algorithm is particularly suitable
for identifying outliers in datasets with a spatial structure, as demonstrated by Birant and Kut (2007), Pavlis et al.
(2018), and Jiang et al. (2019). The points represented in the three-dimensional space of location and cost in
Graph 3 exemplify such a structure.
4.2 Use of DBSCAN in Identifying Out-of-Pattern Transactions (Outliers) at the Manufacturer Stage
The DBSCAN model employs a density-based spatial clustering technique for noise applications, as Uncu et al.
(2006) and Si (2024) discussed. The density of points in a specific region determines the formation of clusters. A
point cannot join a cluster if it does not meet the density or distance criteria, as Birant and Alp (2007) and Hanafi
and Saadatfar (2022) noted. It is essential to highlight that, unlike centroid-based algorithms such as K-Means,
DBSCAN can identify clusters of arbitrary shapes, making it efficient for complex and heterogeneous datasets.
The function that obtains all the neighbors of element p, Epsilon-neighborhood (𝑁𝑝) is defined as follows:
𝑁
𝑝 = 𝑞∈𝐷|𝑑𝑖𝑠𝑡𝑝,𝑞𝜀
(1)
Where: (p) and (q) are points in the dataset 𝐷, 𝑑𝑖𝑠𝑡𝑝, 𝑞 is the distance between (p) and (q), (ε) is the
specified neighborhood radius. A point is considered a core point (Core Point) if the number of points in (𝑁) is
greater than or equal to (MinPts).
As a second definition, we have the concept of directly density-reachable, which states that an element p is
directly density-reachable from an element q if:
𝑝 ∈ 𝑁𝜀𝑞 (2)
|𝑁
𝑝|𝑚𝑖𝑛𝑃𝑇𝑆 (3)
The Border Point is a point that is not a core point but is within the (ε) region of a core point. The Noise Point is
considered noise if it is neither a core point nor a border point. According to Ventorim (2021), the algorithm was
developed to discover clusters and noise from a dataset based on its spatial coverage. DBSCAN uses density to
perform clustering, which enables the algorithm to perform clustering of arbitrary shapes.
The core elements and the border elements are the two types of elements present in a cluster. There are elements
in an ε-neighborhood of a core element, based on its definition, at least minPTS. But this does not apply to
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
59
border elements. Thus, a core element can be directly reachable from a border element; however, the reverse is
not true. A core element has more ε-neighborhood elements than a border element. In the directly
density-reachable criterion, pairs of core elements have symmetry. This means that if p is a core element and q is
also, and if p is directly density-reachable from q, the reverse is also true.
The third concept to be presented uses the previous concept in its definition, given that it is an extension of
directly density-reachable, being called density-reachable and defined as: An element p is density-reachable from
an element q, if there is a chain of elements 𝑝,𝑝
,..., 𝑝
, given that𝑝 = 𝑝 and 𝑝
= 𝑞 such that
𝑝𝑖 1 is directly density-reachable from pi. In the fourth definition, we have the concept of density connected.
An element p is density-connected to an element q if there is an element in which both p and q are
density-reachable. When considering two border elements in a cluster, there is at least one core element from
which both border elements are density reachable.
If there is a dataset D, a cluster C is a non-empty subset of D if it satisfies the following properties:
I. ∀𝑝,𝑞: 𝑆𝑒 𝑝 ∈ 𝐶 𝑒 𝑞 If q is density-reachable from p, then q ∈ C. (Maximality).
II. ∀𝑝, 𝑞 ∈ 𝐶: 𝑝 is density-connected to q. (Connectivity).
This is the density-based cluster concept. The DBSCAN algorithm can divide the dataset into clusters and
identify the members belonging to each cluster, as long as all elements of the same cluster are density-connected.
Figure 5 shows the clustering via DBSCAN. In this figure, the outlier transaction is identified in the
manufacturer's stage. Cluster zero contains 1 observation (cost 2.4).
Figure 5. Clustering via DBSCAN
Source. Own elaboration.
Figure 5 visually represents the clusters identified by the DBSCAN algorithm. These clusters offer valuable
insights into each cluster's geographical distribution and cost values. Cluster zero, which contains only the outlier,
signifies a shipment that warrants deeper scrutiny. This outlier, deviating from the pattern found in the other four
clusters, underscores the DBSCAN algorithm's unique ability to identify elements that do not fit well into any
cluster. The cost scale, ranging from 1 to 10, was designed to accentuate the difference between the average costs
of the 23 shipments and the outlier. The DBSCAN results reveal the formation of five clusters, with one being
the noise point (outlier), further demonstrating the algorithm's proficiency in grouping points based on density.
The Silhouette Coefficient served as a tool to validate the accuracy of the estimation. This coefficient gauges
how well a point has been clustered with other points in the same cluster, considering the proximity of points
within a cluster and the distance to points in the nearest clusters. The K-means algorithm was used as a
benchmark to compare the accuracy of the DBSCAN algorithm. The Silhouette Coefficient value for the
DBSCAN algorithm was 0.711246, while for the K-means algorithm, it was 0.5749625. The higher Silhouette
value indicates that the DBSCAN algorithm outperformed K-means in effective clustering, as evidenced by the
analysis of the Silhouette Coefficient values. A higher Silhouette Coefficient for DBSCAN suggests that points
within each cluster are closer to each other, while points in different clusters are relatively farther apart,
indicating that the clusters are separated from each other.
Notably, including more variables can significantly enrich cluster analysis, especially in contexts such as
customs dispatch, where multiple data dimensions can capture the complexity of operations, and the diversity of
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
60
goods involved. Adding more variables allows for a more nuanced understanding of patterns and relationships in
the data, which can lead to more precise identification of clusters and outliers and provide deeper insights into
international trade behavior. Additional variables might include trade volume, which can indicate risk, especially
for shipments that do not match usual patterns for specific products or trade routes. Transaction frequency can
also indicate risk, with less frequent operators representing a higher risk. Compliance history, including
information on previous violations or irregularities, can provide further insights. The destination country may
present varying levels of risk based on economic, political, and regulatory factors. Another additional
information could be the classification according to exporter frequency, which includes newcomer companies,
continuous and discontinuous companies, and dropouts. These are some of the variables that can significantly
enrich cluster analysis.
Clustering algorithms, as essential tools for enhancing risk assessment processes, automatically group data points
based on inherent similarities. This allows customs authorities to detect unusual patterns that may indicate
fraudulent or non-compliant activities, Steenari, e Nurminen(2023). By analyzing large volumes of transaction
data, clustering algorithms uncover hidden structures and relationships that might not be immediately apparent
through manual inspection. This efficient process reassures customs officials that they are equipped with the best
tools for the job. Consequently, clustering algorithms facilitate the efficient processing of data and empower
customs officials to proactively identify and address potential risks (Galindo González,2024), thereby
safeguarding national security and economic interests.
5. Logistics Agent Risk Assessment Stage
The logistics (or distribution) agent's space is formed by four dimensions that indicate the agent's identification
number, the HS code of the merchandise, the state, and the commercial volume. These parameters (coordinates)
aim to show whether the choice of a particular agent for the shipment in question is appropriate. Additionally,
they indicate whether the merchandise can be shipped through the declared port. It is possible to analyze the
relationship between two of the mentioned parameters (for example, commercial volume and the agent's
experience with the HS code), assuming an export dataset of (i) green coffee, roasted coffee, soluble coffee, and
coffee extracts, (ii) beef (including offal, preparations, and preserves), and (iii) paper and pulp. The data are for
exports via land from Rio Grande do Sul, Brazil. In Graph 5, it can be observed that it is possible to group land
shipments that are out of the region's standard or even the average value associated with the merchandise in
question (beef, paper, and pulp, or green coffee, roasted coffee, soluble coffee, and coffee extracts). In this stage,
a score (risky or safe) can be created for each shipment generated in the blockchain with information such as
those in the Model Insertion Form (Table 2).
The methodology for risk classification in this study involves a systematic approach to assess the risk of logistics
agents based on historical export data. The first step is to calculate the mean (𝜇) and standard deviation (𝜎) of
historical export volumes (𝑋) for each product (i). This provides a baseline for understanding typical export
volumes. Next, for each shipment volume (𝑉
), we compute the deviation from the mean (Eq. 4). This deviation
metric quantifies how much a particular shipment volume deviates from the historical average. To provide a
normalized measure of risk, we define the risk score ((𝑅) based on the deviation from the mean, normalized by
the standard deviation (Eg. 5). This risk score helps identify shipments that significantly deviate from the norm.
A predefined threshold (e.g., two standard deviations) is used to classify shipments as “Risky” in equation 6.
This threshold-based classification ensures that only those shipments with substantial deviations are flagged as
risky.
Deviation from Historical Average:
For each product (i), calculate the mean 𝜇 and standard deviation (𝜎) of historical export volumes (𝑋).
Compute the deviation from the mean for each shipment volume
𝑉
:Deviation =|𝑉
−𝜇
| (4)
Risk Score Calculation:
Define the risk score (𝑅) based on the deviation from the mean, normalized by the standard deviation:
𝑅=||
(5)
Threshold for Risk Classification:
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
61
Classify shipments as "Risky" if the risk score (𝑅) exceeds a predefined threshold (e.g., 2 standard deviations):
Risky if 𝑅>2 (6)
The use of deviation from the mean and standard deviation as metrics for risk assessment is well-established in
various fields, including finance, quality control, and supply chain management. For more information, see
Alexander et al. (2005), Montgomery (2009), Silver and Thomas (2016), Chopra and Meindl (2015), and
Hubbard (2020).
Figure 6. Previous land shipments – Monte Carlo simulation for the volume and specific product shipped from
Rio Grande do Sul
Source: Own elaboration based on Funcex data.
The risk score for evaluating exports results from an analysis that combines the historical behavior of exported
volumes and their relationship with the general market. Initially, the study considers how much the exported
values deviate from the product's historical average (Figure 6). Values significantly exceeding this average,
specifically by more than ±2 standard deviations, classify as potentially risky. A significant percentage variation,
such as growth exceeding 50% compared to the average of the last six months, signals risk. This quantitative
method effectively detects anomalies that may indicate anything from unusual seasonal variations to potential
irregularities in export operations. Furthermore, the analysis delves deeper by considering specific product trends
and seasonality, as well as the growth of the exported value relative to the overall market performance. Products
with clear seasonal patterns that show abnormally high volumes outside these periods are flagged as risky. The
long-term trend of export data helps discern whether an increase in values forms part of a sustainable growth
trajectory or an unexpected spike. Cluster analysis groups shipments by similar characteristics, identifying those
significantly deviating from established commercial norms. Through these analyses, a comprehensive
risk-scoring model is developed that synthesizes these various factors, allowing for the classification of
shipments into risk categories. Figure 5 displays the simulation of 1,000 values for each product using Monte
Carlo Simulation. Values significantly deviating from the mean (e.g., ±2 standard deviations) appear as risky.
Additionally, a significant percentage change compared to the previous period (e.g., more than 50% growth
relative to the average of the last six months) signals risk. The risk-scoring model for exports is designed to be
adaptable, incorporating additional criteria beyond those already mentioned, such as deviation from the mean,
percentage variation, trend analysis, seasonality, market growth comparison, and cluster analysis. Other
dimensions that add value to the model include Regulatory Compliance and Compliance History. An essential
aspect of international trade involves adhering to global and local regulations. An assessment of the regulatory
compliance of exports and the exporter's compliance history will also provide relevant risk information. For
example, exports made by companies with a history of irregularities or regulatory violations appear riskier. This
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
62
criterion considers the frequency and severity of past non-compliances, adjusting the risk score to indicate the
likelihood of regulatory or compliance issues. Another essential aspect to believe in in the Commercial Risk and
Market Diversification analysis is the stability and diversity of commercial relationships. Exports concentrated in
a single market or client may be more susceptible to economic or political variations. Additionally, a sudden
change in commercial relationship patterns, such as unexpected entry into a new market or the abrupt
termination of long-standing commercial relationships without a clear explanation, indicates risk. Incorporating
an analysis of market diversification and the stability of commercial relationships into the risk-scoring model
helps detect risks related to excessive dependence or instability in commercial relationships.
5. Classification Stage
The classification stage is initiated when the transactions from the manufacturer and the transport agent are sent
to the blockchain (Raja Santhi & Muthuswamy, 2022). The submission of the commercial invoice details by the
manufacturer triggers the manufacturer's risk assessment stage. In contrast, the transport agent's transaction
submission to the blockchain triggers the transport agent's risk assessment stage. This article primarily focuses
on presenting a classification method to mitigate the risk associated with these two agents involved in the
customs process. After each risk assessment, the result is promptly sent to the customs administration through
the blockchain system, providing customs with a real-time shipment assessment during the international trade
supply chain stages.
For each stage, the risk of a point (shipment) is obtained by analyzing its location in the "risky" and "non-risky"
subspaces. The evaluation process for each subspace (producer, transport, and logistics agent) determines
whether the new point can be categorized as a normal point or an outlier (risky).
A continuous improvement and feedback loop is established to ensure the ongoing effectiveness of the risk
classification process. The performance of the risk classification models is regularly monitored and adjusted
based on new data and feedback from customs authorities. Input from all stakeholders, including manufacturers,
transport agents, and customs authorities, is collected to identify areas for improvement. By implementing a
robust and dynamic risk classification process, customs authorities can enhance the efficiency and security of the
international trade supply chain. This is achieved by ensuring that shipments are more accurately assessed,
reducing the risk of delays due to unnecessary inspections, and appropriately managed, thereby minimizing the
potential for security breaches.
As mentioned above, the classification criteria rely on quantitative and qualitative factors. Regarding
Quantitative Factors, for example, the Declared Value involves comparing the declared value of goods with
historical data and market trends to identify discrepancies. The Shipping Route evaluation also analyzes the
shipping route for unusual patterns or deviations from standard practices. Furthermore, the Volume and Weight
assessment evaluates the shipment's volume and weight against historical averages and industry standards.
Regarding Qualitative Factors, for example, the Compliance History involves evaluating the compliance history
of the manufacturer and the transport agent, including any past violations or irregularities. The Country Risk
considers the risk associated with the origin and destination countries, including political stability, regulatory
environment, and economic conditions. Additionally, the Product Type assessment evaluates the risk based on
the type of product being shipped, with higher scrutiny for sensitive or high-value goods.
There are numerous possibilities for evaluating the risk of goods in international trade. The process is adaptable
to the objectives of the involved economic agent, whether it be customs, the financing bank, the buyer, or even
the agency providing insurance for the exported cargo. Your role in this process is crucial and significant. Your
unique objectives and expertise are integral to the risk evaluation, making it a collaborative and comprehensive
process.
The classification criteria rely on quantitative and qualitative factors. Table 4 presents Some Possible Criteria
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
63
Table 4. Quantitative and Qualitative Risk Factors
Quantitative Factors Description
Declared Value Comparing the declared value of goods with historical data
and market trends to identify discrepancies.
Shipping Route Analyzing the shipping route for any unusual patterns or
deviations from standard practices.
Volume and Weight Evaluating the shipment's volume and weight against
historical averages and industry standards.
Qualitative Factors Description
Compliance History
Evaluating the compliance history of the manufacturer and
the transport agent, including any past violations or
irregularities.
Country Risk
I am considering the risks associated with the origin and
destination countries, including political stability,
regulatory environment, and economic conditions.
Product Type
We are evaluating the risk based on the type of product
being shipped, with higher scrutiny for sensitive or
high-value goods.
Source. Own elaboration.
6. Integration of the Proposed Method with Blockchain
The proposed method for risk assessment in customs is significantly enhanced by its integration with blockchain
(Nguyen et al., 2022). This integration revolutionizes the efficiency, transparency, and reliability of the risk
assessment process. Blockchain's secure data storage capabilities provide an immutable environment for storing
shipment transaction data, including manufacturer information, shipping agent, invoice details, and importer
transaction history. This immutability ensures the reliability of the information used in the risk assessment, as the
data cannot be manipulated or tampered with.
Blockchain enhances product traceability by tracking how shipments move along the supply chain, from the
manufacturer to the final recipient. This improvement in traceability helps identify potentially risky shipments
and facilitates the investigation of fraudulent activities. Blockchain can automate the risk assessment process by
intelligently executing the steps of the method suggested in the previous sections (Dzhaparov, 2020). This
reduces the need for manual intervention, speeds up the process, and reduces the likelihood of human errors.
Furthermore, the process becomes more transparent and collaborative. Blockchain promotes transparency by
allowing all parties involved in the supply chain, such as manufacturers, shipping agents, customs authorities,
and importers, to access and view transaction data. This transparency helps stakeholders work together and
makes them feel included and part of a team, united in the fight against trade fraud.
Integrating blockchain with the proposed risk assessment method enhances data security and privacy.
Blockchain's decentralized nature ensures that no single entity controls the entire data set, reducing the risk of
data breaches and unauthorized access. Each transaction is encrypted and linked to the previous one, creating a
secure chain of information that is highly resistant to hacking and cyber-attacks. This robust security framework
instills trust and confidence among stakeholders, encouraging them to share accurate and complete data, which
further improves the effectiveness of the risk assessment process.
Integrating the proposed method with blockchain technology extends beyond security and transparency, opening
new avenues for innovation and operational efficiency. Blockchain eliminates the need for intermediaries and
enables process automation through smart contracts, significantly reducing operational costs and processing
times, thereby ensuring financial security and efficiency. Smart contracts, self-executing codes operating within
the blockchain, can automate document verification, goods release, and payment execution, ensuring all parties
fulfill their obligations accurately and promptly, McKinney et al. (2017). This automation accelerates processes
and minimizes the risk of human errors and fraud, creating a more secure and reliable commercial environment.
Blockchain's ability to provide an immutable and verifiable record of all transactions is a game-changer in
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
64
modernizing customs practices. It facilitates auditing and regulatory compliance, allowing for more efficient and
precise oversight of foreign trade operations. This improves governance and compliance and strengthens trust
between trade partners and regulatory authorities. Implementing this technology sets a new standard of
excellence and innovation in international trade, empowering customs administrations to tackle the challenges of
the 21st century with greater agility and effectiveness.
Some countries have already integrated blockchain technology into international trade. Singapore launched the
TradeTrust (Note 2) project, which uses blockchain to digitize and authenticate commercial documents such as
invoices and certificates of origin. This helps reduce fraud and speeds up the customs clearance process, bringing
a new level of efficiency to international trade. TradeTrust was developed to meet the requirements of the
UNCITRAL Model Law on Electronic Transferable Records (MLETR), which was adopted into Singapore
legislation - the Electronic Transactions Act (ETA) in 2021. TradeTrust enables adopters to quickly implement
ETRs as an electronic Bill of Lading that meets the requirements of the MLETR, Singapore ETA, UK Electronic
Trade Documents Act (ETDA), and US (New York and Delaware) laws, and, therefore, they are legally valid
across multiple platforms and systems. TradeTrust's open-source code is freely available. It can be easily
integrated into any enterprise and solution provider's systems to create and verify documents supporting viable
use cases.
The Korea Customs Service (KCS) (Note 3) has launched an initiative to use blockchain to digitize customs
documents, improve the traceability of goods, reduce fraud, and enhance operational efficiency. Korea Customs
has been enhancing its global customs network and striving to create a mutually beneficial trade environment
while offering traders the latest news and trends. A dedicated team has been organized and dispatched to support
our traders further to minimize clearance errors and assist businesses in clearance disputes. This team is a
testament to our commitment to providing comprehensive support. Additionally, Korea Customs is actively
working with the World Customs Organization to secure trade facilitation and security, conducting ODA projects
and delivering various programs through cooperation funds. This includes modernization projects for customs
administration and e-clearance systems in developing countries and capacity-building activities for customs
officers to contribute to mutual growth and advancement of the global society.
7. Final Considerations
The Fourth Industrial Revolution has shown the revolutionary effects of technologies such as Artificial
Intelligence (AI), blockchain, and big data on the world. To remain competitive in an increasingly demanding,
connected, and dynamic market, companies and experts must adapt to the changes brought about by these
innovations. The transition to this new technological era requires more than just technical improvements. It
necessitates a strategic approach that values continuous learning, reinvention, and collaboration among people.
Customs, which play an essential role in international trade, face the challenge of properly managing and
overseeing the entry and exit of goods, requiring the analysis of a large volume of data and collaboration with
various governmental and private organizations.
Amidst the challenges customs activities face, blockchain technology emerges as a pivotal solution, promising to
enhance efficiency and security in international trade. By providing a secure and immutable data storage
environment, blockchain facilitates shipment tracking, fosters transparency in commercial operations, and
automates risk assessment. This not only aids in preventing fraud and irregularities but also improves data
management and collaboration among international trade agents. The integration of blockchain into the customs
process not only optimizes operations but also opens doors to new business opportunities and innovative service
models, underscoring the importance of adaptation and continuous innovation in the face of global trade
challenges.
This study introduces a blockchain-based risk classification method that offers significant advantages to customs
risk management and international trade security. With this method, shipments can be evaluated in real-time at
the manufacturer and carrier stages, leveraging securely and immutably stored data on the blockchain. This not
only enhances the efficiency of the customs procedure but also boosts the transparency and trust of commercial
operations. The automation of risk assessment, made possible by blockchain technology, reduces the need for
manual intervention, thereby minimizing the possibility of human errors and accelerating the customs clearance
process.
The logistics agent's risk analysis, incorporating multiple dimensions such as the agent's identification number,
the HS code of the goods, the state, and the trade volume, provides a comprehensive view of the agent's
suitability for the shipment in question. This multidimensional approach ensures a more accurate assessment of
the risk associated with each shipment, facilitating the identification of potentially risky operations and informed
decision-making by customs authorities. The Monte Carlo simulation, applied to estimate the volume and
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
65
specific product shipped from the State of Rio Grande do Sul, exemplifies the practical applicability of this
method in identifying out-of-pattern shipments, contributing to the effectiveness of the risk scoring model.
Nonetheless, developing a risk-scoring model that synthesizes factors such as deviation from the mean,
percentage variation, trend analysis, seasonality, and comparison with market growth offers a robust tool for
classifying shipments into risk categories. Including additional criteria, such as regulatory compliance and
compliance history, as well as the analysis of market diversification and the stability of commercial relationships,
further enriches the analysis, allowing for a more holistic and informed risk assessment.
This article presented a risk-scoring model for Brazilian exports, considering economic and financial criteria.
The model aimed to provide a tool to support customs agents' decision-making. Notably, the suggested model
can be helpful in various agents related to foreign trade, not only customs agents but also exporters, banks,
insurers, government agencies, and international organizations. The model can assist in risk assessment
objectively and transparently, reducing information asymmetry and transaction costs. The methodologies
presented in this article establish a starting point for a more in-depth and refined analysis of export risks. The
model can be enhanced with more criteria, data, and methods, depending on the availability and quality of
information.
Integrating the proposed risk assessment method with blockchain can transform risk management in international
trade, offering customs authorities a powerful tool for identifying and mitigating risks efficiently and
transparently. This study contributes to the body of knowledge in customs security and international trade,
highlighting the crucial role of blockchain technology in facilitating safer and more regulated trade.
This method can improve processing times, operational costs, transparency, and supply chain security. The
blockchain-based approach effectively identified fraudulent activities, such as value manipulation and
undeclared goods, thereby enhancing compliance and trust among trade partners. However, the study
encountered data availability and quality limitations and the need to adapt the method for various products and
trade routes. Future research should apply this technique to other industries and commercial contexts.
Implementing this method can significantly reduce operational costs and processing times in international trade
while boosting security and transparency. This approach benefits customs authorities and private operators,
fostering a more reliable and efficient commercial environment.
One major limitation involves the availability and quality of data used in the risk analysis model. Inconsistent or
incomplete data can undermine the accuracy of analyses and the effectiveness of the risk classification system.
To address this issue, implementing rigorous data verification and validation protocols before inserting
information into the blockchain proves essential, ensuring the use of only accurate and complete data.
Additionally, adapting the method for various products and trade routes may require machine learning algorithms
and risk classification criteria adjustments, as each context presents unique characteristics and challenges.
Developing specific modules for each product type or route can overcome this limitation, allowing for system
customization that accounts for the particularities of each scenario. Finally, resistance to adopting new
technologies among some stakeholders may pose an obstacle. Thus, it becomes crucial to promote training
sessions and workshops demonstrating the proposed method's benefits and efficacy, facilitating its
implementation and acceptance within the international trade sector.
Acknowledgments
The author thanks the Proficiency program at the State University of Rio de Janeiro (UERJ) and
PROAP-CAPES.
Informed consent
Obtained.
Ethics approval
The Publication Ethics Committee of the Canadian Center of Science and Education.
The journal and publisher adhere to the Core Practices established by the Committee on Publication Ethics
(COPE).
Provenance and peer review
Not commissioned; externally double-blind peer reviewed.
Data availability statement
The data that support the findings of this study are available on request from the corresponding author. The data
are not publicly available due to privacy or ethical restrictions.
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
66
Data sharing statement
No additional data are available.
Open access
This is an open-access article distributed under the terms and conditions of the Creative Commons Attribution
license (http://creativecommons.org/licenses/by/4.0/).
Copyrights
Copyright for this article is retained by the author(s), with first publication rights granted to the journal.
References
Abdisa, T., Getu, M., & Etana, M. (2023). Review on Sanitary and Phytosanitary Agreement. J Vet Heal Sci, 4(2),
53-63.
Alexander, J. M., Frey, R., Embrechts, P., Duffie, D., & Schaefer, S. (2005). Quantitative Risk Management:
Concepts, Techniques and Tools. In Princeton University Press: Princeton and Oxford (p. 538).
Azcárraga, A. A. P., & San Juan, R. (2022). Strengthening the Core Customs Processes through Integrated Risk
Management. Strengthening Customs Administration in a Changing World, 131.
Bhosale, K., Akbarabbas, K., Deepak, J., & Sankhe, A. (2019). Blockchain based secure data storage.
International Research Journal of Engineering and Technology (IRJET), 6(3), 5058-5061.
Birant, D., & Kut, A. (2007). ST-DBSCAN: An algorithm for clustering spatial–temporal data. Data &
Knowledge Engineering, 60(1), 208-221. https://doi.org/10.1016/j.datak.2006.01.013
Berk, RA, & Bleich, J. (2013). Procedimentos estatísticos para previsão de comportamento criminoso: Uma
avaliação comparativa. Criminology & Pub. Pol'y, 12, 513. https://doi.org/10.1177/0263276411417430
Casado-Vara, R., Prieto, J., De la Prieta, F., & Corchado, J. M. (2018). How blockchain improves the supply
chain: Case study alimentary supply chain. Procedia Computer Science, 134, 393-398.
https://doi.org/10.1016/j.procs.2018.07.193
Calixto, J. (2019). Blockchain e sua utilização nos processos logísticos. *FATEC LOG*, São Paulo, 1-10.
Chopra, S., & Meindl, P. (2010). Supply chain management, strategy, planning and operation. Pearson. Upper
Saddle River, NJ, USA.
Dangsawang, B., & Nuchitprasitchai, S. (2024). A machine learning approach for detecting customs fraud
through unstructured data analysis in social media. Decision Analytics Journal, 10, 100408.
https://doi.org/10.1016/j.dajour.2024.100408
Deshpande, A. (2012). Supply chain management dimensions, supply chain performance and organizational
performance: An integrated framework. International Journal of Business and Management, 7(8), 2.
http://dx.doi.org/10.5539/ijbm.v7n8p2
Dzhaparov, P. (2020). Application of blockchain and artificial intelligence in bank risk management. Икономика
и управление, 17(1), 43-57.
Dos Santos, D. R. (2023). Finanças Digitais. Tokens e o Potencial da Tecnologia Blockchain. ISBN
978-6500-65942-9. ASIN: B0BZWK5455. Amazon.
Drobot, E. V., Klevleeva, A. R., Afonin, P. N., & Gamidullaev, S. N. (2017). Risk management in customs
control. Ekonomika Regiona = Economy of Regions, (2), 550-558.
Dujak, D., & Sajter, D. (2019). Blockchain applications in supply chain. SMART Supply Network, 21-46.
https://doi.org/10.1007/978-3-319-91668-2_2
Fahmi, A. I. (2018). Risk Identification in International Trade by Matching Company and Product Classification
Data.
Galindo González, D. R. (2024). Leveraging spatial data to enrich analytics systems (Master's thesis, Universitat
Politècnica de Catalunya).
Gao, W., Hatcher, WG, & Yu, W. (2018, julho). Uma pesquisa sobre blockchain: Técnicas, aplicações e desafios.
Em 2018, 27ª conferência internacional sobre comunicação e redes de computadores (ICCCN) (pp. 1-11).
IEEE. https://doi.org/10.1109/ICCCN.2018.8487348
Hanafi, N., & Saadatfar, H. (2022). A fast DBSCAN algorithm for big data based on efficient density calculation.
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
67
Expert Systems with Applications, 203, 117501. https://doi.org/10.1016/j.eswa.2022.117501
Haque, E. U., Shah, A., Iqbal, J., Ullah, S. S., Alroobaea, R., & Hussain, S. (2024). A scalable blockchain based
framework for efficient IoT data management using lightweight consensus. Scientific Reports, 14(1), 7841.
https://doi.org/10.1038/s41598-024-58578-7
Haile, T. T. (2024). Web's Progression: Moving from Passive Content Consumption to Active Content Creation
and Content Validation. International Journal of Business and Management, 18(6), 135-135.
https://doi.org/10.5539/ijbm.v18n6p135
Howse, R., Eliason, A., & Trebilcock, M. (2005). The regulation of international trade. Routledge.
https://doi.org/10.4324/9780203799796
Hubbard, D. W. (2020). The failure of risk management: Why it's broken and how to fix it. John Wiley & Sons.
Huria, A. (2019). Facilitating Trade and Logistics for E-Commerce: Building Blocks, Challenges, and Ways
Forward. World Bank.
Jiang, X., Ma, J., Jiang, J., & Guo, X. (2019). Robust feature matching using spatial clustering with heavy
outliers. IEEE Transactions on Image Processing, 29, 736-746. https://doi.org/10.1109/TIP.2019.2934572
Juma, H., Shaalan, K., & Kamel, I. (2019, December). Customs-based distributed risk assessment method. In
International Symposium on Parallel Architectures, Algorithms and Programming (pp. 417-429). Singapore:
Springer Singapore. https://doi.org/10.1007/978-981-15-2767-8_37
Karklina-Admine, S., Cevers, A., Kovalenko, A., & Auzins, A. (2024). Challenges for Customs Risk
Management Today: A Literature Review. Journal of Risk and Financial Management, 17(8), 321.
https://doi.org/10.3390/jrfm17080321
Lima, B., Hitomi, F., & Oliveira, G. (2018). Aplicação da tecnologia blockchain em ambientes corporativos.
Fasci-Tech, 1(13), 1-8.
McKinney, S. A., Landy, R., & Wilka, R. (2017). Contratos inteligentes, blockchain e a próxima fronteira do
direito transacional. Wash. JL Tech. & Arts, 13, 313.
Montgomery, D. C. (2019). Introdução ao controle estatístico de qualidade. John Wiley & Sons.
Nikoofal, M. E., Pourakbar, M., & Gumus, M. (2023). Securing containerized supply chain through public and
private partnership. Production and Operations Management, 32(7), 2341-2361.
https://doi.org/10.1111/poms.13980
Nguyen, S., Chen, P. S. L., & Du, Y. (2022). Avaliação de risco de sistemas integrados de blockchain para
transporte marítimo de contêineres: Uma análise de cenários de múltiplos eventos. Transportation Research
Part E: Logistics and Transportation Review, 163, 102764.
Okazaki, Y. (2017). Implicações de big data para alfândegas - Como pode dar suporte a capacidades de
gerenciamento de risco. WCO Research Paper, 39.
Pavlis, M., Dolega, L., & Singleton, A. (2018). A modified DBSCAN clustering method to estimate retail center
extent. Geographical Analysis, 50(2), 141-161. https://doi.org/10.1111/gean.12138
Rizzardi, A., Sicari, S., Miorandi, D., & Coen-Porisini, A. (2022). Securing the access control policies to the
Internet of Things resources through permissioned blockchain. Concurrency and Computation: Practice and
Experience, 34(15), e6934. https://doi.org/10.1002/cpe.6934
Raja Santhi, A., & Muthuswamy, P. (2022). Influência da tecnologia blockchain na cadeia de suprimentos de
manufatura e logística. Logística, 6(1), 15.
Rouhani, S., & Deters, R. (2021). Estrutura de confiança de dados usando tecnologia blockchain e validação de
transação adaptativa. IEEE Access, 9, 90379-90391. https://doi.org/10.1109/ACCESS.2021.3091327
Si, Y. (2024). A Novel Density-Based Spatial Clustering of Application with Noise Method for Data Clustering.
IJLAI Transactions on Science and Engineering, 2(3), 1-8.
Silver, E. A., Pyke, D. F., & Thomas, D. J. (2016). Inventory and production management in supply chains. CRC
Press. https://doi.org/10.1201/9781315374406
Steenari, J., & Nurminen, J. K. (2023). Geospatial DBSCAN Hyperparameter Optimization with a Novel
Genetic Algorithm Method.
Tacconi, L., Boscolo, M., & Brack, D. (2003). National and international policies to control illegal forest
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
68
activities. Royal Institute of International Affairs.
Tariq, M. U. (2024). Revolutionizing Health Data Management with Blockchain Technology: Enhancing
Security and Efficiency in a Digital Era. In Emerging Technologies for Health Literacy and Medical
Practice (pp. 153-175). IGI Global. https://doi.org/10.4018/979-8-3693-1214-8.ch008
Uncu, O., Gruver, W. A., Kotak, D. B., Sabaz, D., Alibhai, Z., & Ng, C. (2006, October). GRIDBSCAN: GRId
density-based spatial clustering of applications with noise. In 2006 IEEE International Conference on
Systems, Man and Cybernetics (Vol. 4, pp. 2976-2981). IEEE. https://doi.org/10.1109/ICSMC.2006.384571
Ventorim, I. D. M. (2021). BIRCHSCAN: Um método de aproximação do DBSCAN para grandes conjuntos de
dados.
Wang, M., Wu, Y., Chen, B., & Evans, M. (2020). Blockchain and supply chain management: a new paradigm
for supply chain integration and collaboration. Operations and Supply Chain Management: An International
Journal, 14(1), 111-122. http://doi.org/10.31387/oscm0440290
Notes
Note 1. Due to the General Data Protection Law (LGPD - Lei Geral de Proteção de Dados, Law No.
13.709), since 2018, the federal government no longer discloses results (sales volume) by company.
Note 2. https://www.imda.gov.sg/how-we-can-help/international-trade-and-logistics/tradetrust
Note 3. https://www.customs.go.kr/kcs/main.do
Appendix A - Coding of the Model Sheet for Blockchain Insertion (JSON, JavaScript Object Notation)
Figure 2 - Pseudo Code for Coding of the Model Sheet for Blockchain Insertion
{
"Order_ID": "123456789",
"Order_Date": "2023-09-15",
"Manufacturer_Location": {
"Address": "Rua Exemplo 123, Zona Industrial",
"City": "São Paulo",
"Country": "Brazil"
},
"Invoice_Details": {
"Invoice_Number": "FAT-001234",
"Date": "2023-09-15",
"Items": [
{
"Description": "Product A",
"Quantity": 100,
"Unit_Price": 10.00
},
{
"Description": "Product B",
"Quantity": 200,
"Unit_Price": 5.00
}
],
"Total_Value": 2000.00
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
69
},
"Country_of_Origin": "Brazil",
"Product_Description": [
{
"Product": "Product A",
"Specifications": "Technical specifications of Product A",
"HS_Code": "123456"
},
{
"Product": "Product B",
"Specifications": "Technical specifications of Product B",
"HS_Code": "789012"
}
],
"Declared_Value": 2000.00,
"Shipping_Terms": {
"Incoterms": "FOB",
"Carrier": "XYZ Transport Company"
},
"Estimated_Arrival_Date": "2023-10-05",
"Associated_Documentation": {
"Certificates_of_Origin": ["Certificate123.pdf"],
"Licenses": ["ExportLicense456.pdf"]
}
}
Source: Own elaboration
Once prepared, this JSON can be sent to a blockchain system, where it will be stored immutably. Each field
within the JSON serves a specific purpose, ensuring that all necessary information for risk assessment,
compliance, and traceability is present and easily accessible (Figure 2).
Appendix 2 – Pseudo Code for Data Modeling via DBSCAN for Identifying Possible Outliers
Figure 3 shows the pseudo-code for data modeling via DBSCAN to identify possible outliers. The R-Project,
specifically the dbscan() function available in the dbscan package, was used.
Figure 3 - Pseudo code for data modeling via DBSCAN for identifying possible outliers
# Load the dbscan package
if (!requireNamespace("dbscan", quietly = TRUE)) install.packages("dbscan")
library(dbscan)
data <- data.frame(
longitude = c(.) ,
latitude = c(.) ,
cost = c(.) ,
HScode = rep("SH0201", 24),
ijbm.ccsenet.org International Journal of Business and Management Vol. 19, No. 6; 2024
70
city = c(.)
)
# Data normalization
data_norm <- scale(data)
# Apply DBSCAN to the normalized data
set.seed(123) # For reproducibility
dbscan_result <- dbscan(data_norm, eps = "", minPts = "")
# View the results
print(dbscan_result)
Source. Own elaboration.
The pseudo-code performs the following steps: For each point in the dataset, the number of neighbors within a
distance (𝜀) is calculated. A point is then marked as a core point if it has a number of neighbors greater than or
equal to (MinPts). Points that are not core put are within the distance (𝜀) of a core point are marked as border
points. Points that are neither core nor border points are considered outliers. Therefore, to detect the outlier point
through the distance between regular points, you would configure DBSCAN with an appropriate value of (𝜀)and
(MinPts) based on the spatial distribution of your dataset. The point termed an outlier, having a significantly
different (lower) cost than other points and possibly being isolated in three-dimensional space (considering
longitude, latitude, and price), would be identified as an outlier by the algorithm due to its low neighbor density.
Copyrights
Copyright for this article is retained by the author(s), with first publication rights granted to the journal.
This is an open-access article distributed under the terms and conditions of the Creative Commons Attribution
license (http://creativecommons.org/licenses/by/4.0/).
