There are numerous implications of cheap, micro-scale and computationally-sophisticated sensors, ranging in scope from national security to neural prosthetics and implantable health monitoring devices. Unfortunately, the capabilities of mobile and autonomous sensors are hindered today by their limited power budgets. The focus of the Analog Lab is to develop hardware implementations of signal processing that are orders of magnitude more power- and area-efficient than is achievable with current techniques. To this end, we exploit advances in analog subthreshold design, floating gate technology, and tools from dynamical systems theory. Also, our designs are heavily inspired by systems and processes that are observed in nature and in biology. Our current work is in smart sensors for computer audition, a single-chip solution for electroencephalography, analog signal processing for next-generation image sensors and noise-suppression algorithms for hearing prosthetics.