Autonomous following of heavy vehicles has the potential to greatly reduce the costs of operating platoons of trucks. Computer controlled following allows a vehicle to follow another vehicle at close range, increasing the gas mileage of trucks within the platoon and alleviating some problems related to driver fatigue. A visual system developed by UCLA, IRIS-2 (Infrared Ranging via Image Subtraction), is installed on the following vehicle and is used to obtain the relative position and relative yaw to the preceding vehicle. The autonomous following problem must be defined properly for analysis and design of control algorithms.

The heavy vehicle model used is similar to the magnetometer-based control model, except that the road-reference frame is not practical for following. The unsprung mass model (origin at tractor CG) is used for the autonomous following model, but two vehicles are necessary for the problem definition. The relative lateral displacement and yaw between vehicles are modeled in terms of the states of the following vehicle, and the effect of motion by the preceding vehicle is treated as a disturbance to these relative parameters.

The problem of tracking a vehicle's trajectory using IRIS-2 without inter-vehicle communication is investigated. Maximizing the ability of a single vehicle to track the trajectory of the preceding vehicle will minimize the propagation of errors throughout a platoon of trucks. Two techniques are proposed for tracking the preceding vehicle's trajectory. The first method ensures that the following vehicle forward velocity vector is always pointing toward the rear of the preceding trailer. The second method assumes that the following vehicle will travel at constant yaw rate and longitudinal velocity over the truck spacing distance and calculates the projected lateral error of this circular trajectory at the truck spacing distance. Simple, manual loop-shaping controllers are designed to test each proposed tracking method, and simulations are performed.