Creating information systems that sense and respond to the physical environment is a complex activity, requiring technical skills from disparate areas of practice, such as computer programming and electronic circuitry. Although recent tools have lowered barriers to creating such systems, they tend to be too technical and constraining for creating systems to be a feasible everyday activity. These tools often rely on traditional interaction techniques and draw makers’ attention away from the system being built, thereby limiting makers’ physical movement, removing systems from their use context, and preventing contextualized experimentation with system designs. This thesis explores techniques for designing tools with support for making systems a more feasible everyday activity. I present the novel design and evaluation of such a tool called Pixel designed to let makers use intuitive knowledge derived from experience with the physical world, rather than technical expertise, in creating custom information systems in the course of everyday life.