Home / Current Issue / Paper 1724441
Multi-Agent Task Automation System
Subject area: Science,Engineering and Technology · Area of research: Artificial Intelligence and Multi-Agent Systems
DOI: 10.64388/IREV10I4-1724441
Abstract
Multi-Agent Systems (MAS) offer an effective approach for solving complex and large-scale problems in distributed artificial intelligence by dividing them into smaller subtasks handled by multiple autonomous agents. Each agent can perceive its environment, process information, make decisions, and perform actions according to its assigned objectives. Agents communicate, collaborate, and coordinate to achieve individual and collective goals, providing advantages such as scalability, adaptability, flexibility, robustness, and fault tolerance. The proposed Multi-Agent Task Automation System uses heterogeneous agents dynamically assigned tasks according to their capabilities, resource availability, current workload, task priority, execution cost, and contextual relevance. A mediator-based coordination mechanism supports agent discovery and task routing while allowing direct peer-to-peer communication when necessary. Dynamic task allocation and load balancing help distribute work efficiently. Reinforcement learning can enhance agent decision-making by allowing agents to learn from feedback and improve their strategies over time. The system also incorporates fault detection, recovery, authentication, secure communication, and trust evaluation to improve reliability and security. The implementation uses Python, AutoGen, large language models, APIs, and external tools, with structured message passing enabling agents to interact and execute tasks collaboratively. Tasks are initiated through user input and processed using iterative communication, reasoning, and tool execution. The proposed approach demonstrates the potential of MAS to improve task execution efficiency, response time, resource utilization, adaptability, scalability, and fault tolerance compared with traditional centralized approaches. However, communication overhead, coordination complexity, security risks, synchronization, and efficient task allocation remain important challenges.
Keywords
multi-agent systems (MAS); autonomous agents; distributed artificial intelligence; agent communication; coordination and collaboration; task allocation; AutoGen; large language models; reinforcement learning; scalability; adaptability; fault tolerance; intelligent decision making
References
[1] Dorri A, Kanhere SS, Jurdak R. Multi-agent systems: A survey. IEEE Access. 2018;6:28573–28593.
[2] Russell S, Norvig P. Artificial intelligence: A modern approach. 3rd ed. Upper Saddle River (NJ): Prentice Hall; 2010.
[3] Wooldridge M. An introduction to multiagent systems. 2nd ed. Wiley; 2009.
[4] Weiss G. Multiagent systems: A modern approach to distributed artificial intelligence. MIT Press; 1999.
[5] Jennings NR. An agent-based approach for building complex software systems. Communications of the ACM. 2001;44(4):35–41.
[6] Lesser VR. Cooperative multiagent systems: A personal view of the state of the art. IEEE Transactions on Knowledge and Data Engineering. 1999;11(1):133–142.
[7] Wooldridge M, Jennings NR. Intelligent agents: Theory and practice. The Knowledge Engineering Review. 1995;10(2):115–152.
[8] Wu Q, et al. AutoGen: Enabling next-gen LLM applications via multi-agent conversation. arXiv:2308.08155 [Preprint]. 2023. https://arxiv.org/abs/2308.08155
How to cite this paper
@article{1724441,
author = {Mohammed Lutfi, Raju S., Marilingan, Mariam Fathima, Nanda Kumar},
title = {Multi-Agent Task Automation System},
journal = {Iconic Research And Engineering Journals},
year = {2026},
volume = {10},
number = {4},
pages = {4511-4519},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1724441.pdf},
abstract = {Multi-Agent Systems (MAS) offer an effective approach for solving complex and large-scale problems in distributed artificial intelligence by dividing them into smaller subtasks handled by multiple autonomous agents. Each agent can perceive its environment, process information, make decisions, and perform actions according to its assigned objectives. Agents communicate, collaborate, and coordinate to achieve individual and collective goals, providing advantages such as scalability, adaptability, flexibility, robustness, and fault tolerance. The proposed Multi-Agent Task Automation System uses heterogeneous agents dynamically assigned tasks according to their capabilities, resource availability, current workload, task priority, execution cost, and contextual relevance. A mediator-based coordination mechanism supports agent discovery and task routing while allowing direct peer-to-peer communication when necessary. Dynamic task allocation and load balancing help distribute work efficiently. Reinforcement learning can enhance agent decision-making by allowing agents to learn from feedback and improve their strategies over time. The system also incorporates fault detection, recovery, authentication, secure communication, and trust evaluation to improve reliability and security. The implementation uses Python, AutoGen, large language models, APIs, and external tools, with structured message passing enabling agents to interact and execute tasks collaboratively. Tasks are initiated through user input and processed using iterative communication, reasoning, and tool execution. The proposed approach demonstrates the potential of MAS to improve task execution efficiency, response time, resource utilization, adaptability, scalability, and fault tolerance compared with traditional centralized approaches. However, communication overhead, coordination complexity, security risks, synchronization, and efficient task allocation remain important challenges.},
keywords = {multi-agent systems (MAS); autonomous agents; distributed artificial intelligence; agent communication; coordination and collaboration; task allocation; AutoGen; large language models; reinforcement learning; scalability; adaptability; fault tolerance; intelligent decision making},
month = {October},
doi = {https://doi.org/10.64388/IREV10I4-1724441}
}