The Visual Programming Challenge (VPC) is a competition designed to test the effectiveness of visual/multimodal programming languages at programming certain problems. The objective of the challenge is to focus research efforts on visual/multimodal programming approaches that offer the greatest potential of allowing `ordinary` people to `program` tasks. One of the problems of unfocused research is that the emphasis tends towards novel, rather than effective. The intent of the VPC is to promote research that leads to effective solutions. The particular problem domain of the competition requires real-time control of an intelligent robot vehicle within an unknown system of roads composed of standard Lego road tiles. The solution to this problem requires managing a representation for a 2D data structure (of unknown dimensions), constructing closures over dynamic data structures (to determine if all paths connect), planning (to find the best route to an incomplete path needing further exploration), and real-time control (many of the robot interactions must occur between specific operation states). The results of visual/multimodal programming have already had an impact on a number of related fields, including scientific visualization and human-computer interfaces. This impact will grow as emphasis is increased on bringing more people into daily use of computers for both business and personal interests. This challenge serves to help focus research in visual/multimodal programming languages and thereby to improve the effectiveness of this technology for everyday use.
