Enrique H. Ruspini
AI Center
SRI International

Intelligent Highway-Vehicle Control


The development of effective controllers and planners for intelligent highway systems requires control of a variety of motion and perceptual actions and the smooth integration of behaviors seeking multiple, possibly conflicting, objectives. For example, automated cruise control behaviors must be integrated with emergency handling maneuvers (e.g., tire blowup) and with collision-avoidance techniques.

It is also desirable that existing low-level regulators, which have been proven successful to control specific tasks, be integrated and managed by higher-level controllers that mediate their interactions and assign operational priorities on the basis of situation-dependent considerations. Furthermore, this integration must be broad enough to encompass also the regulation of sensors and the integration of multimodal sensor capabilities.

The fuzzy-logic approach to intelligent control developed by SRI is, therefore, specially suited to develop vehicle-based controllers for intelligent highway systems. Under contract with the National Advanced Highway Systems Consortium, SRI is has developed an encompassing concept for the integration of various perception techniques, notably GPS and image understanding, sensor-fusion methods, and reactive/purposive maneuver regulation on the basis of an architecture based on this control approach [Galijan et al 1996].

| Enrique H. Ruspini | Artificial Intelligence Center | SRI International |
Last updated on March 18, 1999