Predictive Displays and Latency Protections for High Latency Teleoperation

We will develop and demonstrate robust and reliable methods for high-latency teleoperation with Predictive Display (PD), Latency Protection (LP) and Supervisory Control (SC). PD conveys the effects of latent commands not yet reflected in the display. Augmented Reality PD (ARPD) overlays an icon indicating the predicted position and orientation using nominal terrain and motion models. Virtual Reality PD (VRPD) warps the image to the predicted perspective using a “2½ D” model constructed via video feature tracking. LP limits translation and rotation rates, and shapes acceleration response, as functions of observed latency and operator behavior. In extreme situations, LP brakes to prevent operator overreaction. LP is inspired by AntiLock Braking: it has no observable effect within safety margins. Phase I will implement ARPD and LP on a Packbot, and begin VRPD. The Option will test Visual Servoing SC (VSSC) and open-loop Displacement Command SC (DCSC) for high-latency teleoperation, and implement VRPD. VSSC developed under a prior separate effort. ARPD, LP and DCSC are software-only solutions that can be implemented on existing UGVs without additional hardware. We will conduct developmental testing to optimize and compare the methods based on task performance, stress-related biosignals and subjective workload assessment.