An Abstract Behavior Representation for Robust, Dynamic Sequencing in a Hybrid Architecture
Endeavors in mobile robotics focus on developing autonomous vehicles that operate in dynamic and uncertain environments. By reducing the need for human-in-the-loop control, unmanned vehicles are utilized to achieve tasks considered dull or dangerous by humans. Because unexpected latency can adversely affect the quality of an autonomous system’s operations, which in turn can affect lives and property in the real-world, their ability to detect and handle external events is paramount to providing safe and dependable operation. Behavior-based systems form the basis of autonomous control for many robots. This thesis presents the unified behavior framework, a new and novel approach which incorporates the critical ideas and concepts of the existing reactive controllers in an effort to simplify development without locking the system developer into using any single behavior system. The modular design of the framework is based on modern software engineering principles and only specifies a functional interface for components, leaving the implementation details to the developers. In addition to its use of industry standard techniques in the design of reactive controllers, the unified behavior framework guarantees the responsiveness of routines that are critical to the vehicle’s safe operation by allowing individual behaviors to be scheduled by a real-time process controller. The experiments in this thesis demonstrate the ability of the framework to: 1) interchange behavioral components during execution to generate various global behavior attributes; 2) apply genetic programming techniques to automate the discovery of effective structures for a domain that are up to 122 percent better than those crafted by an expert; and 3) leverage real-time scheduling technologies to guarantee the responsiveness of time critical routines regardless of the system’s computational load.