Event

Neuromorphic decision dynamics enable deadlock-free navigation

1 reports · 1 sources

Himani Sinhmar, Vaibhav Srivastava, and Naomi Ehrich Leonard propose neuromorphic decision dynamics (NDD), a decentralized, event-driven controller for safe navigation in crowded environments. The authors claim that NDD requires no communication, no preset priorities, no neighbor goal models, and no estimation of group size or global network properties; agents self-organize into local conflict groups and form passage orders by coupling internal decisions and attention states to physical states and the relative motion of local observations. The authors derive explicit conditions on sensing range, feedback gains, response time scales, and actuation limits, under which they claim that finite heterogeneous agent populations of arbitrary and variable size can avoid collisions and form passage orders in finite time without falling into deadlock or livelock. Large-scale simulations, comparisons with existing methods, and hardware experiments are used to demonstrate NDD's performance in complex crowded scenarios and to validate its scalability.