Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Aug 27;14(17):3004.
doi: 10.3390/foods14173004.

Generative AI and Blockchain-Integrated Multi-Agent Framework for Resilient and Sustainable Fruit Cold-Chain Logistics

Abhirup Khanna  1 , Sapna Jain  2 , Anushree Sah  1 , Sarishma Dangi  3 , Abhishek Sharma  3 , Sew Sun Tiang  4 , Chin Hong Wong  5   6 , Wei Hong Lim  4
Affiliations

Generative AI and Blockchain-Integrated Multi-Agent Framework for Resilient and Sustainable Fruit Cold-Chain Logistics

Abhirup Khanna et al. Foods. .

Abstract

The cold-chain supply of perishable fruits continues to face challenges such as fuel wastage, fragmented stakeholder coordination, and limited real-time adaptability. Traditional solutions, based on static routing and centralized control, fall short in addressing the dynamic, distributed, and secure demands of modern food supply chains. This study presents a novel end-to-end architecture that integrates multi-agent reinforcement learning (MARL), blockchain technology, and generative artificial intelligence. The system features large language model (LLM)-mediated negotiation for inter-enterprise coordination, Pareto-based reward optimization balancing spoilage, energy consumption, delivery time, and climate and emission impact. Smart contracts and Non-Fungible Token (NFT)-based traceability are deployed over a private Ethereum blockchain to ensure compliance, trust, and decentralized governance. Modular agents-trained using centralized training with decentralized execution (CTDE)-handle routing, temperature regulation, spoilage prediction, inventory, and delivery scheduling. Generative AI simulates demand variability and disruption scenarios to strengthen resilient infrastructure. Experiments demonstrate up to 50% reduction in spoilage, 35% energy savings, and 25% lower emissions. The system also cuts travel time by 30% and improves delivery reliability and fruit quality. This work offers a scalable, intelligent, and sustainable supply chain framework, especially suitable for resource-constrained or intermittently connected environments, laying the foundation for future-ready food logistics systems.

Keywords: blockchain; cold-chain logistics; generative AI; multi-agent reinforcement learning; sustainable food systems.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflicts of interest.

Figures

Figure 1
Figure 1
SLA compliance.
Figure 2
Figure 2
Temperature control precision.
Figure 3
Figure 3
Fuel consumption vs. travel time.
Figure 4
Figure 4
Emissions vs. Travel Distance.
Figure 5
Figure 5
Delay Probability vs. Traffic State.
Figure 6
Figure 6
Reduction in fuel consumption over iterations.
Figure 7
Figure 7
Reduction in travel time.
Figure 8
Figure 8
Reduction in emissions.
Figure 9
Figure 9
Delivery rate improvement.
Figure 10
Figure 10
State Evolution: Temperature.
Figure 11
Figure 11
State Evolution: Humidity.
Figure 12
Figure 12
Action impact.
Figure 13
Figure 13
Predicted spoilage probability: oranges.
Figure 14
Figure 14
Predicted spoilage probability: strawberries.
Figure 15
Figure 15
Predicted spoilage probability: bananas.
Figure 16
Figure 16
Delivery speed adjustments: oranges.
Figure 17
Figure 17
Delivery speed adjustments: strawberries.
Figure 18
Figure 18
Delivery speed adjustments: bananas.
Figure 19
Figure 19
Spoilage reduction.
Figure 20
Figure 20
Inventory mismatch.
Figure 21
Figure 21
Shelflife optimization.
Figure 22
Figure 22
Vitamin C retention.
Figure 23
Figure 23
SLA violation count.
Figure 24
Figure 24
Adaptive energy usage.
Figure 25
Figure 25
RL model comparison.
Figure 26
Figure 26
Module contribution.
Figure 27
Figure 27
Pareto frontier.
Figure 28
Figure 28
SLA transaction latency.
Figure 29
Figure 29
KDE plot: temperature.
Figure 30
Figure 30
KDE plot: humidity.
Figure 31
Figure 31
Comparison of spoilage rate.
Figure 32
Figure 32
Comparison of SLA violations.
Figure 33
Figure 33
Comparison of energy consumption.
Figure 34
Figure 34
Spoilage rate distributions.
Figure 35
Figure 35
Energy consumption distribution.
Figure 36
Figure 36
Modular agent responsiveness.
Figure 37
Figure 37
Agent performance under varying network uptime.
Figure 38
Figure 38
Impact of power outages on temperature control and spoilage.
Figure 39
Figure 39
Performance metrics under adverse shipment conditions.
Figure 40
Figure 40
Correlation: energy consumption and carbon emissions.
See this image and copyright information in PMC

References

    1. Barat S., Kumar P., Gajrani M., Khadilkar H., Meisheri H., Baniwal V., Kulkarni V. International Workshop on Multi-Agent Systems and Agent-Based Simulation. Springer International Publishing; Cham, Switzerland: 2019. Reinforcement learning of supply chain control policy using closed loop multi-agent simulation; pp. 26–38.
    1. Yang Y., Wang M., Wang J., Li P., Zhou M. Multi-Agent Deep Reinforcement Learning for Integrated Demand Forecasting and Inventory Optimization in Sensor-Enabled Retail Supply Chains. Sensors. 2025;25:2428. doi: 10.3390/s25082428. - DOI - PMC - PubMed
    1. Zeng M., Wu Y., Xing X., Tang W., Xu H. Coordinating principal–agent and incentive strategy of cold chain logistics service in fresh food supply chain. PLoS ONE. 2024;19:e0306976. doi: 10.1371/journal.pone.0306976. - DOI - PMC - PubMed
    1. Liu H., Zhang J., Zhou Z., Dai Y., Qin L. A Deep Reinforcement Learning-Based Algorithm for Multi-Objective Agricultural Site Selection and Logistics Optimization Problem. Appl. Sci. 2024;14:8479. doi: 10.3390/app14188479. - DOI
    1. Lau H., Tsang Y.P., Nakandala D., Lee C.K. Risk quantification in cold chain management: A federated learning-enabled multi-criteria decision-making methodology. Ind. Manag. Data Syst. 2021;121:1684–1703. doi: 10.1108/IMDS-04-2020-0199. - DOI

LinkOut - more resources