Streaming Video Modeling for Robotics Teleoperation

摘要

Most of the current fielded Unmanned Ground Vehicle (UGV) functionality is dependent on the ability to drive the UGV using tele-operation technology. In addition, a large number of payloads require tele-operation to perform the mission function. Tele-operation technology is dependent on providing the operator streaming video, which is reliant on radio capabilities along with video format, resolution and compression routines. There have been Army efforts to perform real-time network modeling as part of Program Executive Office-Integration (PEO-I). These are primarily related to the passing of C2 tactical information from vehicle to vehicle. Ground Vehicle Robotics (GVR) has funded a proof of principle effort that culminated in a demonstration performed in February, 2011. This effort modeled the impact of latency, packet/data loss and distorted signal on streaming video being sent from a virtual UGV to the Operator Control Unit (OCU). These distorted signals, cause loss of vehicle control. The initial capability will allow for tradeoff analysis of radios, waveforms and antennas in performing tele-operation and payload control of manipulator arms in various rural and urban environments. This paper will discuss the tradeoffs, architecture and design of the system in support of the February, 2011 demonstration along with potential follow-on activities and capabilities.