Innovation Talk at TARDEC by Dr. Tulga Ersal

Dr. Ersal, photo credit to University of MichiganRecently, Dr. Tulga Ersal, Assistant Research Scientist from the University of Michigan, Ann Arbor, gave an Innovation Seminar talk titled "Driving Unmanned Ground Vehicles at High Speeds: From Teleoperation to Full Autonomy" at U.S. Army TARDEC in Warren, MI.

The TARDEC Innovation Seminar is a weekly event where TARDECians hear about new reasearch, technologies and approaches to solving Army problems. The talks may be given by researchers from within TARDEC or from outside institutions.

Abstract: This talk will highlight ongoing research efforts in the Automotive Research Center to address the problems of teleoperation and fully autonomous operation of large Unmanned Ground Vehicles (UGVs) at high speed.
        Unmanned ground vehicles (UGVs) are gaining importance in military applications because of their significant potential to increase mission performance, combat effectiveness, and personnel safety. They span a wide spectrum in their mode of operation ranging from teleoperated, in which the remote human operator controls all actions of the UGV, to fully autonomous, where the UGV is completely controlled by the computer.
        Communication delays negatively affect mobility performance in teleoperation. As delays increase, controlling the UGV becomes more challenging, and this negative impact is exacerbated at higher speeds. The first half of the talk will present our research efforts to improve mobility performance in teleoperation by compensating delays in a vehicle- and operator-agnostic way to enable a robust and easily implementable solution.
        High speeds also present a challenge to fully autonomous operation with respect to obstacle avoidance in unknown and unstructured environments, because, in addition to avoiding the obstacles, vehicle dynamic limits become an important safety concern. Therefore, the second half of the talk will highlight our research efforts to improve autonomous mobility by developing an algorithm that can perform high-speed obstacle avoidance at the dynamic limits of the UGV in unknown and unstructured environments.