FUNCTION ALLOCATION METRICS APPLIED TO MULTI-AGENT AEROSPACE EXAMPLE

摘要

This paper reviews eight function allocation issues identified in the literature. Based on these issues, the paper describes a series of metrics which are designed to be generated by computational models, observations in high-fidelity human-in-the-loop simulations or real operations. This paper provides an example of a subset of the metrics to show how they can be used to demonstrate the differences in function allocation in multi-agent teams in the aviation domain. Specifically, the metrics are operationalized into the WMC simulation framework which is briefly described. The results illustrate that the metrics are able to detect differences between the function allocations and that the individual function allocations do not perform uniformly well across all metrics. While the function allocation method is illustrated in the context of an aviation example, the principles, metrics, modeling approach are broadly applicable across domains and should have relevance to the design and evaluation of new forms of automation for process control applications.