Pervasive sensing and computing technology will play a key role in future buildings. In the US, for example, federal sustainability goals mandate that 50 percent of commercial buildings must become net-zero energy by 2050. To realize this goal, existing buildings will require a wide range of retrofits and improvements. However, to benchmark buildings and verify the intended effects of the improvements, many critical building parameters, including electricity, gas, and water usage, must be sensed and analyzed both before and after the improvements are made. Furthermore, many of the improvements themselves—such as advanced daylight harvesting, where indoor lighting is adjusted in a finegrained manner in response to outdoor solar irradiance and the degree of shading—require their own sensors that can monitor occupancy, light level, glare, and shade setting. However, most sensors installed on walls, work surfaces, ceilings, shades, and many other places can’t be plugged into mains power, and users won’t tolerate frequent battery replacement. This suggests that a different and more scalable sensing paradigm is required. Motivated by these emerging applications and our own difficulties in fielding long-lived indoor sensors at modest scale, we decided to explore the factors preventing broader adoption of pervasive sensing and computing within buildings.