Our mobile robot, Spinoza, embodies a sophisticated real-time vision system for control of a responsive mobile robot. Balancing the real-time constraints of a robot in a dynamic environment challenges the limits of both technology and our scientific understanding of embedded systems. Dynamic environments are unpredictable, asynchronous, and require a low latency in response, while visual information processing require high data-rate communications and significant computation.
Spinoza, as seen in Figure 1, is a self-contained robot, with host support. It consists of a Real World Interface (RWI) B-12 base with an RGB (colour) camera, mounted on a Directed Perception pan-tilt platform, on top, and trinocular monochrome stereo cameras in the body.
Figure 1: Spinoza: the visually guided mobile robot
To provide a context for the design issues involved in this system, we begin by describing previous work in our lab on developing vision-based robotics systems that are antecedents of Spinoza. Also we present our research goals and a description the tasks Spinoza is to perform. Section 2 describes how the functional requirements of Spinoza and the development environment shapes the the choice of computational architecture and communication protocols. Section 3 describes the robot hardware. Section 4 describes in detail the workings of all the components of the robot from the hardware to the software. We finally present the experimental results and conclude with a discussion about the future of mobile robotics.