Intelligent Two-Sector Sensor for Use in an Autonomous Follower Vehicle, 09-R9603

Printer Friendly Version

Principal Investigators
Norman E. Carmichael
Rod Draney
Mark R. Gildersleeve

Inclusive Dates:  01/01/06 – 01/22/07

Background - The military has demonstrated leader-follower ground vehicles that function well in most conditions, but at a multi-million dollar modification cost per vehicle. Many systems use an "electronic bread crumb" approach that can require two Global Positioning System (GPS) receivers, a radio frequency (RF) link from one unit to the next, a vision system for the manned leader to monitor follower vehicles, and extensive collision avoidance systems. The lead vehicle constructs a trail using GPS coordinates as it travels. It then broadcasts these coordinates to a following vehicle, which moves to each point using a point-to-point navigation system. Other approaches have included inertial navigation, pre-programmed mapped path, and path improvement or modifying the infrastructure. Obstacle recognition and avoidance, path recognition and tracking, and failure recovery (e.g., lack of satellite communications) results in additional complexity. These "high-end" approaches that offer reasonable robustness are not cost feasible for standard truck convoys.

Approach - The project had three main objectives: develop a proof-of-concept for a low-cost intelligent sensing system for use in an autonomous follower vehicle that can determine the angle and distance to a lead vehicle; review hardware and firmware solutions for use in the proof-of-concept; and compare the operational limits of the system with military convoy vehicles requirements.

Accomplishments - A proof-of-concept based on existing commercial off-the-shelf radio frequency infrared sensors was developed as was a simulation package capable of modeling various configurations of autonomous vehicle convoys. Simulations of technologies for implementing a leader-follower were also developed.

The simulations indicate that a combined RF/IR sensor capable of determining distance and angle to a specific lead vehicle is a viable option when combined with GPS or IMU. This solution relies on a proposed RF sensor.

The 3D Path and Terrain Viewer allows the operator to visually evaluate a convoy path.

2007 Program Home