A Semi-Autonomous Control Method to Improve Performance of Small Unmanned Ground Vehicles With Communication Latency

Abstract

Teleoperated unmanned ground vehicles are very useful in environments that are hazardous for humans. When controlled manually, speed of operation can be very slow due to degraded and delayed feedback of information to/from the vehicle’s environment. Adding autonomy to the vehicle can make control for the human teleoperator easier and improve performance. This paper presents a semi-autonomous control method for avoiding collisions while driving a vehicle. The method is well suited for small unmanned ground vehicles in unstructured environments (i.e. environments without predefined roads/paths to follow). The semi-autonomous control method and the effect of communication latency are evaluated with a human subject study (N = 20) involving teleoperation of a simulated robot search task. Results show that while semi-autonomy does improve performance at low communication latency, the improvement is much larger at higher latencies.