Canonical instabilities of autonomous vehicle systems

Rodrick Wallace (1)

1 New York State Psychiatric Institute

Formal argument suggests that command, communication and control systems can remain stable in the sense of the Data Rate Theorem that mandates the minimum rate of control information required to stabilize inherently unstable 'plants', but may nonetheless, under fog-of-war demands, collapse into dysfunctional modes at variance with their fundamental mission. We apply the theory to autonomous ground vehicles under intelligent traffic control in which swarms of interacting, self-driving devices are inherently unstable as a consequence of the basic irregularity of the road network. It appears that such 'V2V/V2I' systems will experience large-scale failures analogous to the vast propagating fronts of power network blackouts, and possibly less benign, but more subtle patterns of 'psychopathology' at various scales.

Mots clés: driving experience; groupoid; information theory; turbulence; vehicle density; control theory

Domaines: Intelligence artificielle [cs.AI]

Rodrick Wallace. Canonical instabilities of autonomous vehicle systems. 2016. ⟨hal-01275653⟩