29 May 2023
Homogeneity is often considered to be a defining characteristic in the swarm paradigm. In simulation, physical homogeneity can be imposed on the system. In a real-world setting however, heterogeneity arises naturally from mechanical differences in the sensors and actuators of the robots. In particular, faults will inevitably occur over the course of system run-time limiting the capabilities of certain individuals and giving rise to a level of heterogeneity.
In this work, EMERGE partners from the University of Bristol present a decentralized, data-driven method of fault detection where individual robots self-detect faults based on measures of their local environment – either in the range of sensing or on-board. The authors summarize results for this method and discuss how it may be useful to measure diversity in a heterogeneous system - can metrics based on local sensing discriminate between a range of given states and can we build a model for the real-time evaluation of diversity in the system?
Read the paper in the link below.
