Presentation Details
Adaptive Agent Community Framework for Real-Time Autonomous Decision Ecosystems
ID：150
View Protection：ATTENDEE
Updated Time：2026-07-27 12:58:35 Hits：7
Online
Start Time：2026-07-31 14:55（Asia/Kolkata）
Duration：15min
Session：S7 Disruptive Technologies for Manufacturing »
Abstract
Abstract—In autonomously operating decision ecosystems, more and more intelligent agents will be deployed in a distributed way, and the agents have to cooperate and operate under conditions where limited bandwidth and changing environmental conditions occur. Centralized controllers, federated learning pipelines and swarm heuristics solve some aspects of the problem, but are all difficult to simultaneously optimize for accuracy, trust, latency & resilience at scale. This paper introduces the Adaptive Agent Community Framework (AACF), which is a self-organising architecture in which autonomous agents participate in communities that are dynamically formed according to the agents' trust estimates, share knowledge among themselves, and achieve adaptive consensus on whatever they have to do using a learning decision engine that constantly receives feedback from the execution of the decision. The conceptual structure combines the process of trust assessment with the creation of the community and also links it with adaptive learning in a closed operational concept and not a static pipeline. Mathematical models are introduced to determine trust scoring, community affinity, knowledge utility, decision confidence, weighted consensus, community stability and adaptation efficiency and a twelve-step operational algorithm is presented. AACF achieves an accuracy in decision making of 95.1%, decisiveness in governance of 92.6%, average response latency after adoption of about 96ms, outperforming the Centralized, Federated Multi-agent and Swarm Intelligence baselines, highlighting the level of accuracy, robustness, scalability and adaptivity, in addition to reducing the communication overhead by up to 33% while applying Simulation experiments to different numbers of agents from 10 to 100. From these results, it can be concluded that trust-aware adaptation with community participation can be a sound direction towards achieving scalable approaches to Real-Time Autonomous Decision Ecosystems.
Keywords
adaptive agent community; multi-agent systems; trust assessment; consensus generation; autonomous decision-making; real-time systems; distributed intelligence.
Speaker
Elamathi E
ASSISTANT PROFESSOR K.Ramakrishnan College of EngineeringSubmission Author
All Comments
Important Date
-
Conference Date
Jul 30
2026
toAug 01
2026
-
Jul 28 2026
Registration deadline
-
Jul 30 2026
Draft paper submission deadline
Sponsored By
The United Societies of Science
Organized By
Kongunadu College of Engineering and Technology
Supported By
IEEE Section
IEEE Madras Section
IEEE Madras Section
Contact Information
Previous Conferences
-
2026-07-30 India Trichy
2026 International Conference on Networks Computers and Communications
Submit Comment