Systems using networks of embedded hardware devices have the potential to be used for a variety of applications in several disciplines, from control networks to robotics research. Each device performs some function and collaborates with the other devices to accomplish a task. In order to build such systems, developers are required to draw on knowledge from a variety of fields, such as circuit design and analysis, embedded programming, networking, and software engineering. Developers with all of these skills are rare, and the time required for others to learn the required skills slows development time and increases costs.
If these hardware devices could be viewed as software objects, then programmers could apply their current skills and design methodologies to the development of embedded device networks, reducing the problem to the application of object-oriented software engineering principles. This thesis deals with the problem of defining the requirements of a software object, the parallelism between a hardware and a software object, and the implementation of a hardware/software framework for building embedded device networks using intelligent hardware devices called Physical Object Devices, or PODs.