Traditional Artificial Intelligence applications have always been outside the world of embedded systems, with AI languages and programming techniques better suited to the lab instead of the circuit board. Times and technology have changed. Quantum leaps in embedded processor horsepower, web connectivity, and cheap and abundant on-board memory are breaking down the barriers for comparatively intelligent embedded applications. This paper explains some of the nuts and bolts concepts and techniques behind an embedded approach to AI programming. Algorithms are supplied in most of the major sections.