Patrick Chwalek

Paper: (in process)

Summary: Distributed indoor sensing offers rich insight into the dynamic interior states of the built environment. However, individually powering a network of sensors is difficult and oftentimes leaves designers in a constant battle of trading off system complexity, size, and sensor sampling frequency to increase the battery life of their systems. For wireless power options, there has been considerable work in harvesting power through various untethered means (e.g., thermoelectric, radio-frequency, etc.) but photovoltaic (PV) options are usually the most suitable for their simplicity, practicality, and superior power density. The main drawbacks of using photovoltaics is that if incident light is intermittent, the system needs to have enough overhead energy storage to stay active during dark periods. In indoor environments, this issues is exacerbated since PV-powered systems can be placed in closed office spaces with limited access to daylight so their main method of harvesting energy is from the luminaires. Given frequently changing occupancy patterns, indoor light source availability is highly variable so creating a secondary system to aid in wirelessly powering PV-based sensor systems is warranted to keep them from going offline. In Spotlight, we seek to create a motorized focused light source that can scan the room and learn the optimal light points to charge each node within its field-of-view. This system will enable us to create reliable untethered PV-based systems to collect robust sensory data for accurate indoor environment modelling and actuation. Further details will be released in our upcoming paper.