eFlux: Simple Automatic Adaptation for Environmentally Powered Devices

Energy management is a critical problem in designing mobile computing systems, especially when those systems depend on harvesting energy from environmental sources, such as solar or wind. Environmental sources are highly variable and difficult to predict, which is often complicated further by device mobility. In this demo, we present a simple approach for developing energy-aware applications using a high-level data flow oriented coordination language. This language, eFlux, is an extension of the Flux (Burns et al., 2006) coordination language, which provides a simple interface for specifying an energy adaptation policy, which can then be implemented automatically by the underlying runtime system. This approach allows a system designer to change the underlying adaptation algorithms (e.g. energy source prediction) without modifying the application. Also, the data flow programming style of Flux simplifies program profiling and performance prediction. In this demo, we will present our experience, to date, using eFlux, including both working system and simulation results. We will also demonstrate an energy-aware GPS tracking device for tracking threatened Wood Turtles in Western Massachusetts