Title: Exploration of robust and intelligent navigation algorithms to ensure off-road autonomous vehicle mobility

Authors: Michael Cole; Kumar B. Kulkarni; Jordan Ewing; Seth Tau; Chris Goodin; Paramsothy Jayakumar

Addresses: US Army DEVCOM Ground Vehicle Systems Center, 6501 E 11 Mile Rd., Warren, Michigan, USA ' US Army DEVCOM Ground Vehicle Systems Center, 6501 E 11 Mile Rd., Warren, Michigan, USA ' Michigan Technological University, 1400 Townsend Dr, Houghton, MI 49931, USA ' US Army DEVCOM Ground Vehicle Systems Center, 6501 E 11 Mile Rd., Warren, Michigan, USA ' Center for Advanced Vehicular Systems, Mississippi State University, 10 Lee Blvd., Mississippi State, MS 39762, USA ' US Army DEVCOM Ground Vehicle Systems Center, 6501 E 11 Mile Rd., Warren, Michigan, USA

Abstract: The combat capabilities development command (DEVCOM) ground vehicle systems centre (GVSC) is supporting unmanned ground vehicle (UGV) development. Past experimentations of a military UGV demonstrated that its autonomous mode performed worse than the tele-operated mode. To address this, a systematic investigation into path planners for military vehicles in off-road environments was executed. A UGV simulator was used to evaluate vehicle and planner performance through a range of obstacle avoidance scenarios in deformable soil to capture the effects of vehicle-terrain interactions across multiple soil types. Monte Carlo methods were used to evaluate the robustness of five path planners ranging from classical to state-of-the-art planners, with normally-distributed variability in environmental and vehicle initial conditions. After running thousands of simulations, results show how each algorithm compares to one another in several key metrics including overall success rates. These results will help inform decisions in future military UGV path planner selection.

Keywords: navigation algorithms; vehicle mobility; autonomous vehicles; unmanned ground vehicles; UGV; path planning; vehicle-terrain interactions; VTI.

DOI: 10.1504/IJVP.2024.140004

International Journal of Vehicle Performance, 2024 Vol.10 No.3, pp.239 - 267

Received: 22 May 2023
Accepted: 01 Aug 2023
Published online: 15 Jul 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article