Using Data Confluences in a Distributed Network with Social Monitoring to Identify Fault Conditions

This paper discusses the potential benefits of socially attentive monitoring in multi-agent systems. A multi-agent system with this feature is shown to detect and identify when an individual within the network fails to operate correctly. The system that has been developed is capable of detecting a range of common faults such as stuck at zero by allowing communication between peers within a software agent network. Further adaption to the model allows an improvement in system response without introduction of specific control design algorithms