Integrating AI for data analysis and automated decision-making in logistics smart contracts

1. Introduction

The rapid digitalization of global logistics networks has necessitated the integration of intelligent technologies to address challenges such as information asymmetry, operational inefficiencies, and lack of transparency. Amid these transformations, blockchain-based smart contracts have emerged as pivotal instruments for enhancing trust and automation in logistics and supply chain management. However, the complexity of logistics operations, characterized by multidimensional data flows and dynamic decision environments, calls for advanced computational methods to support real-time analysis and autonomous decision making. In this context, artificial intelligence (AI) serves as a complementary technology that extends the functional capacities of smart contracts by enabling adaptive analytics, intelligent decision modeling, and predictive capabilities.

Incorporating AI into blockchain-based logistics solutions also addresses the issue of data veracity and decision accountability. AI models, when applied to supply chain analytics, require access to trustworthy and tamper-proof data sources. Blockchain ensures data integrity, provenance, and auditability, thus serving as a reliable foundation for AI-driven analytics (Agrawal et al., 2023). Additionally, when AI agents are tasked with automated decision making, smart contracts can enforce the agreed-upon parameters and provide a decentralized mechanism for executing those decisions without manual intervention. This synergy reduces human error, streamlines multi-party collaborations, and enhances response time to logistics contingencies (Alqarni et al., 2023). Despite the technological promise, the integration of AI with smart contracts in logistics systems introduces several problem domains that must be systematically addressed. Ultimately, the problem setting for the integration of AI applications in data analysis and automated decision making in logistics smart contracts revolves around the convergence of three core technological pillars: data, intelligence, and trust. Data provides the operational substrate, AI imparts the cognitive dimension, and smart contracts operationalize trust and enforcement. However, the realization of this triad in functional logistics systems requires addressing architectural, computational, legal, and organizational challenges through interdisciplinary research and innovation. While blockchain-based smart contracts have already demonstrated value in logistics by enhancing transparency and process automation, their potential is significantly amplified when integrated with AI-driven data analytics and decision-making capabilities. This integration transforms static contracts into adaptive, context-aware systems that support intelligent logistics management. The pressing research problems include AI-contract interoperability, verification of hybrid systems, scalability of AI computations within decentralized environments, and the ethical governance of automated decisions. Addressing these problems is critical for building resilient, efficient, and intelligent logistics networks in an increasingly interconnected global economy.

2. Literature review

The convergence of artificial intelligence (AI), blockchain technology, and smart contracts has transformed the landscape of logistics and supply chain management. This literature review provides an in-depth examination of current academic and industrial research on AI applications in data analysis and automated decision-making within logistics smart contracts. Key themes include the integration of AI with blockchain, the design and execution of intelligent smart contracts, data analytics for real-time logistics decision-making, and privacy and security in decentralized systems.

2.1. Integration of artificial intelligence and blockchain in logistics

Recent studies emphasize the increasing demand for decentralized and intelligent logistics systems capable of operating autonomously and securely. Abdelhamid, Sliman, and Ben Djemaa (2024) explored an AI-enhanced blockchain model tailored for scalable Internet of Things (IoT)- based supply chains. In this model, AI agents interpret real-time data from IoT devices and optimize smart contract actions, allowing for dynamic responsiveness to environmental conditions such as delays, inventory shortages, or transportation bottlenecks. Similarly, Chukwu et al. (2024) proposed an AI-powered framework for enhancing resilience and security in logistics supply chains. The integration enables predictive analytics, anomaly detection, and autonomous decision- making based on real-time logistics data, reinforcing the need for adaptive smart contracts that respond to operational uncertainties without human intervention. This research adopted a questionnaire survey involving 281 managers, with the aim to comprehensively examine the current state of AI integration across the U.S. supply chain sector, with focus on some key components like real-time tracking, cost optimization, and risk management.

Virovets et al. (2024) further advanced this integration by introducing cryptographic oracles, which serve as trusted bridges between AI systems and blockchain-based smart contracts. These oracles facilitate secure, real-time data flows from AI modules to contract logic, ensuring that the smart contracts execute decisions based on verified analytics outcomes.

Figure 1: Traditional interaction of AI with smart contracts via an oracle

Artificial Intelligence (AI) and Distributed Ledger Technology (DLT) together address complex tasks by optimizing and automating business processes and creating innovative new products. Despite their shared digital nature, integrating these two technologies is a challenging process that requires sophisticated solutions. AI relies on large amounts of data and computational power, which are difficult to provide within distributed ledgers. However, the integration of DLT with AI, particularly its interaction with smart contracts, is made possible through the use of an intermediary data exchange and transfer mechanism known as an oracle.

2.2. Smart contracts in supply chain logistics

Smart contracts are self-executing code segments deployed on blockchain platforms to enforce agreements autonomously. Within the logistics domain, these contracts enable process automation, conditional payment settlements, and real-time compliance enforcement. Alqarni et al. (2023) examined the role of blockchain-based smart contracts in streamlining logistics operations. Their findings indicated that by embedding business logic directly into the contract, processes such as cargo tracking, proof of delivery, and customs clearance could be executed automatically. Agrawal et al. (2023) implemented a blockchain-based framework to support collaboration in supply chain networks. The smart contracts within this framework monitored milestones across multiple parties, triggering contractual obligations without manual validation. Such automation fosters transparency and reduces transaction times, which are critical for global logistics. The work by Bottoni et al. (2020) introduced the concept of intelligent smart contracts that leverage AI algorithms to evaluate contextual data before execution. This innovation expands the scope of smart contracts from static rule enforcement to dynamic, data-driven decision engines capable of responding to complex scenarios in supply chains.

2.3. Data analysis and AI-driven decision making

The volume and complexity of logistics data necessitate advanced analytical capabilities. AI techniques such as machine learning (ML), deep learning (DL), and natural language processing (NLP) have demonstrated effectiveness in extracting actionable insights from logistics datasets. Rane, Choudhary, and Rane (2024) provided a comprehensive analysis of how ML models contribute to sustainability and resilience in logistics networks. Predictive analytics support route optimization, demand forecasting, and anomaly detection, which feed into smart contracts for automated adjustments in service-level agreements (SLAs) or delivery conditions. Kumar et al. (2022) presented an architecture for secure smart contracts in cloud-based manufacturing logistics. The system utilizes AI for resource allocation and production scheduling. Based on real-time demand fluctuations and manufacturing constraints, the AI engine feeds optimized decisions into Ethereum-based smart contracts that adjust logistics workflows instantly.

Security remains central in logistics operations where data integrity and contractual reliability are paramount. Sun and Yu (2020) introduced a formal verification framework for analyzing security vulnerabilities in blockchain smart contracts. Their methodology ensures that contracts governing logistics operations remain resistant to attacks, logic flaws, and malicious data injections, which is vital when decisions are automated. Li, Han, and Tang (2021) proposed a privacy-preserving storage scheme for logistics data using blockchain. This approach leverages zero-knowledge proofs and homomorphic encryption to protect sensitive data while enabling smart contracts to execute decisions based on obfuscated datasets. Such methods allow AI to analyze data without compromising confidentiality, a key requirement in inter-organizational logistics collaborations. Balcerzak et al. (2022) discussed the role of smart contracts in decentralized governance systems, highlighting the necessity for transparent, secure, and auditable decision-making. Their framework can be extended to logistics by establishing governance protocols that enforce ethical AI deployment and ensure compliance with cross-border regulatory standards.

2.4. Frameworks and models for logistics automation

Several conceptual and practical models have emerged to support the use of AI in smart contract-based logistics. Groschopf, Dobrovnik, and Herneth (2021) introduced a maturity model for evaluating supply chain sustainability using smart contracts. Their framework assesses whether logistics contracts are capable of capturing ESG (Environmental, Social, Governance) metrics and adapting operational decisions based on sustainability goals.Raja Santhi and Muthuswamy (2022) explored the influence of blockchain on manufacturing logistics, identifying AI-integrated smart contracts as a catalyst for autonomous material handling and delivery scheduling. This integration ensures real-time alignment between production activities and supply chain movements, minimizing downtime and inventory discrepancies.

3. Data and research methods

This study applies a mixed-methods approach, combining quantitative analysis of 20 smart contract transactions from a blockchain-based logistics platform with qualitative interviews from 5 industry experts. Machine learning models, including decision trees and neural networks, were used to analyze performance metrics such as execution time, cost, and error rate. Regression and clustering techniques identified significant correlations between AI integration and logistics efficiency. Thematic analysis of expert interviews revealed key adoption challenges and strategic considerations. This method provides a focused and data-driven understanding of AI’s role in optimizing logistics smart contracts.

Data collection: Data Collection: With 120 transactions extracted from the blockchain system (https://etherscan.io) of a logistics platform using smart contracts. The data includes variables such as: execution time (in seconds), transaction cost (gas fee), error rate (number of

failed transactions), and the level of AI interaction in each transaction. The raw data underwent a pre-processing phase to normalize value ranges and detect outliers. Z-score normalization was applied to execution time and gas cost. Transactions exceeding three standard deviations from the mean were marked as anomalies and excluded from the regression analysis. This resulted in a clean dataset of 112 transactions. AI interaction levels were encoded as ordinal values (0 to 3), allowing for inclusion in quantitative models.

4. Findings and discussion

4.1.  Overview of quantitative data patterns and pre-processing

This research is grounded in the analytical evaluation of 112 smart contract transactions retrieved from a blockchain-based logistics platform, verified via Etherscan. Each transaction included the following variables: execution time (in seconds), transaction cost (measured in gas units), error rate (quantified by failed transactions per batch), and the degree of AI interaction. The degree of AI interaction was categorized into four levels: no AI, low AI (rule-based automation), moderate AI (machine learning-based prediction), and high AI (end-to-end decision automation using neural networks).

Table 1. Summary statistics for transaction variables (n = 112)

4.2. Regression analysis of AI interaction and logistics performance

A multiple linear regression model was constructed to examine the relationships between AI interaction level and logistics performance metrics. The dependent variables were execution time, transaction cost, and error rate. The independent variable of interest was the AI interaction level. Control variables included transaction volume and node latency (data gathered via network logs). Results showed a statistically significant negative correlation between AI interaction and execution time (p < 0.01), suggesting that higher levels of AI support reduced contract execution time. Similarly, transaction cost showed a mild negative association with AI level (p = 0.041), while error rate displayed a strong negative correlation (p < 0.001).

These findings confirm that AI integration in smart contracts plays a decisive role in enhancing operational efficiency. Reduced execution time implies higher throughput in logistics networks, while lower error rates suggest more accurate fulfillment and improved reliability of automated decisions. To gain further insight into how different patterns of AI adoption influence performance, a k-means clustering algorithm was applied to group transactions into clusters based on execution time, cost, error rate, and AI level. The Elbow method determined the optimal number of clusters as three.

Table 3. Cluster characteristics

The third cluster, characterized by high AI usage, demonstrated significant improvements across all performance indicators. Execution times were 49.9% faster than those in Cluster 1. Error rates dropped by over 85%, suggesting that neural-network-based AI plays a crucial role in automated validation and transaction correction.

4.3. Machine learning model performance in predicting transaction outcomes

Two machine learning models were trained to predict transaction outcomes based on input features including AI level, node latency, and transaction volume: a decision tree (CART) and a neural network (MLP). Evaluation of the models used an 80/20 train-test split. Root mean square error (RMSE) and R² values were calculated to assess predictive accuracy.

Table 4. Model performance metrics

While this study reveals significant patterns, certain limitations must be acknowledged. The transaction dataset, though diverse, was limited to a single blockchain platform. Broader validation across multiple platforms could strengthen external validity. Interviewee selection was limited to five experts; expanding the panel across geographies would yield more nuanced perspectives. Future studies should evaluate the lifecycle carbon cost of AI-powered smart contracts, assess real- time collaborative decision-making between human agents and AI, and explore the ethics of algorithmic bias in logistics contract enforcement. Integration with quantum-proof cryptographic protocols may also be a significant research trajectory.

AI applications in logistics smart contracts yield demonstrable benefits in transactional efficiency, cost reduction, and error minimization. Quantitative analysis confirmed statistically significant relationships between AI level and operational performance. Machine learning models, particularly neural networks, provided predictive strength in modeling logistics behavior. Thematic insights from expert interviews contextualized these findings, revealing both strategic potential and systemic challenges. The convergence of AI and blockchain continues to reshape logistics automation. However, the success of AI in autonomous smart contract execution depends on robust data pipelines, legal clarity, and system interoperability. This study contributes to the evolving discourse by offering a comprehensive, data-backed, and stakeholder-informed evaluation of AI’s role in the logistics smart contract ecosystem.

5. Conclusion

The integration of Artificial Intelligence (AI) in logistics smart contracts has demonstrated significant potential to enhance operational efficiency, reduce costs, and improve reliability in decentralized logistics platforms. This study, through a mixed-methods approach, combined the analysis of 112 blockchain-based smart contract transactions with insights from five expert interviews to offer a comprehensive understanding of AI’s role in automated decision-making processes. Quantitative findings revealed that transactions with high levels of AI interaction exhibited faster execution times (mean = 13.5 seconds), lower transaction costs (mean gas fee = 0.00185 ETH), and significantly fewer errors (error rate = 2.5%) compared to those with low AI involvement (mean execution time = 17.2 seconds, gas fee = 0.00221 ETH, error rate = 6.7%). These metrics highlight the tangible benefits of AI in managing the dynamic complexities of logistics operations, particularly in terms of smart contract execution optimization and cost- efficiency. Qualitative insights emphasized that despite these advantages, adoption challenges persist, especially regarding interoperability between AI modules and blockchain environments, algorithm transparency, and trust in AI-generated decisions. Experts noted that the success of AI integration depends heavily on technical literacy within logistics firms, robust governance frameworks, and the standardization of data input formats to facilitate seamless automation. The results affirm that AI-supported smart contracts are not only feasible but also beneficial in streamlining logistics workflows. However, realizing the full potential of this technological synergy requires strategic investments in infrastructure, regulatory clarity, and capacity building across stakeholders in the logistics supply chain.

References:

Abdelhamid M. M., Sliman L., & Ben Djemaa R. (2024). AI-enhanced blockchain for scalable IoT-based supply chain. Logistics, 8(4), 109.

Agrawal T. K., Angelis J., Khilji W. A., Kalaiarasan R., & Wiktorsson M. (2023). Demonstration of a blockchain-based framework using smart contracts for supply chain collaboration. International journal of production research, 61(5), 1497-1516.

Alqarni M. A., Alkatheiri M. S., Chauhdary S. H., & Saleem S. (2023). Use of blockchain- based smart contracts in logistics and supply chains. Electronics, 12(6), 1340.

Arumugam S. S., Umashankar V., Narendra R., Mujumdar A. P., Holler J., & Hernandez A. (2018). IOT enabled smart logistics using smart contracts. In 2018 8th International conference on logistics, informatics and service sciences (LISS) (pp. 1-6). IEEE.

Balcerzak A. P., Nica E., Rogalska E., Poliak M., Klieštik T., & Sabie O. M. (2022). Blockchain technology and smart contracts in decentralized governance systems. Administrative Sciences, 12(3), 96.

Bottoni P., Gessa N., Massa G., Pareschi R., Selim H., & Arcuri E. (2020). Intelligent smart contracts for innovative supply chain management. Frontiers in Blockchain, 3, 535787.

Chukwu N., Yufenyuy S., Ejiofor E., Ekweli D., Ogunleye O., Clement T., ... & Obunadike C. (2024). Resilient chain: AI-enhanced supply chain security and efficiency integration. International Journal of Scientific and Management Research, 7(3), 46-65.

Groschopf W., Dobrovnik M., & Herneth C. (2021). Smart contracts for sustainable supply chain management: Conceptual frameworks for supply chain maturity evaluation and smart contract sustainability assessment. Frontiers in Blockchain, 4, 506436.

Kumar A., Abhishek K., Nerurkar P., Ghalib M. R., Shankar A., & Cheng X. (2022). Secure smart contracts for cloud‐based manufacturing using Ethereum blockchain. Transactions on Emerging Telecommunications Technologies, 33(4), e4129.

Li H., Han D., & Tang M. (2021). A privacy-preserving storage scheme for logistics data with assistance of blockchain. IEEE Internet of Things Journal, 9(6), 4704-4720.

Lu Y., & Cheng C. (2022). Computer Application and Information Security (ICCAIS 2021). In Proc. of SPIE, 12260, 1226001-1.

Raja Santhi A., & Muthuswamy P. (2022). Influence of blockchain technology in manufacturing supply chain and logistics. Logistics, 6(1), 15.

Rane N., Choudhary S., & Rane J. (2024). Artificial intelligence and machine learning for resilient and sustainable logistics and supply chain management. Available at SSRN 4847087.

ỨNG DỤNG AI TRONG PHÂN TÍCH DỮ LIỆU VÀ RA QUYẾT ĐỊNH TỰ ĐỘNG TRONG HỢP ĐỒNG THÔNG MINH LOGISTICS

Bùi Thị Minh Thu

Trường Đại học Kinh tế - Đại học Đà Nẵng

Tóm tắt:

Nghiên cứu này khám phá việc tích hợp Trí tuệ nhân tạo (AI) vào các hợp đồng thông minh trong lĩnh vực logistics dựa trên công nghệ blockchain, nhấn mạnh vai trò của AI trong việc nâng cao khả năng phân tích dữ liệu và tự động hóa quá trình ra quyết định. Kết quả cho thấy các ứng dụng AI có thể cải thiện đáng kể hiệu quả logistics thông qua việc xử lý dữ liệu theo thời gian thực, thực thi hợp đồng linh hoạt và giảm thiểu sai sót do con người. Thông qua phân tích chủ đề các cuộc phỏng vấn chuyên gia, nghiên cứu xác định một số thách thức chính trong quá trình triển khai, bao gồm khả năng tương thích hệ thống, mức độ tin cậy vào quyết định do AI đưa ra và sự thiếu vắng khung dữ liệu tiêu chuẩn hóa. Bằng cách xem xét các yếu tố này, nghiên cứu đóng góp vào diễn đàn học thuật về sự tích hợp giữa blockchain và AI trong lĩnh vực logistics, đồng thời đề xuất một số khuyến nghị chính sách thực tiễn như xây dựng tiêu chuẩn quản trị AI, phát triển cơ chế khuyến khích đổi mới và thiết lập các môi trường thử nghiệm hợp tác nhằm đảm bảo an toàn trong quá trình thử nghiệm và triển khai trên quy mô lớn.

Từ khóa: trí tuệ nhân tạo, hợp đồng thông minh, logistics, tự động hóa.