Complementary Metal-Oxide-Semiconductor (CMOS) integrated circuits can provide high levels of functionality within ultra-compact silicon microchips. Recently, efforts have been made to develop standalone microchips that include all necessary functions required to form a full electronic sensing system, including power scavenging, signal detection, data processing and storage, and wireless communications. This total integration allows for the creation of complete electronic systems, measuring only millimeters or even microns across, that can operate autonomously without physical tethers to any off-chip equipment. The compact form factor of autonomous chips has opened new application areas for which more conventional sensing systems are too bulky, including deployment as computational engines on mobile platforms that have extremely strict size and weight budgets. This thesis will discuss an autonomous chip designed to be mounted on honey bees to perform flight tracking, and another set of chips designed to serve as the brains of walking and swimming microrobots. The pursuit of total integration has introduced unique circuit design challenges that have necessitated novel solutions, including power harvesting using on-chip silicon photovoltaics and the development of optoelectronic signal detection interfaces and other circuits that can tolerate the marginal power resources available from these photovoltaics. Additionally, signal processing algorithms and systems-level modeling techniques that have been developed to predict and improve the performance of these chips are presented.