Binderr LLC develops a new type of smart sensors that deliver task-relevant actionable data. These functional sensors are based on new technology that bypasses the traditional processing pipeline commonly used to go from low-level data (e.g. color, range, orientation, etc.) to an actionable world-system model. Solutions based on Binderr technology have been demonstrated to be orders of magnitude smaller than existing solutions in terms of Size, weight, power, computations, cooling, and cost (SWaP C3), as well as latency. Binderr develops two classes of functional sensors: Direct Functional Sensors and Evolved Functional Sensors. Direct Functional Sensors, employ a single light-weight transform on sensor data (cameras, lidar, radar, etc.) that yields multiple invariant patterns directly specifying functional properties and events of an activity in the world (e.g., field of safe travel). One example in this class is Binderr’s Cyclo™, a small footprint collision detection and braking solution using a single video camera. Another is DirectScene6™, a video camera based scene detector solution providing object detection, localization, and distances for ground vehicles/robots. Evolved Functional Sensors are created through a new machine learning technique (i.e. a new class of Dynamical Artificial Neural Networks) to evolve radically small solutions. For example RoamBot™, an evolved digital circuit of 50 transistors capable of autonomous collision-free navigation, homing to visible charging stations, and responding to verbal commands, using a video camera, IR emitters, and a microphone. Another evolved solution is Bindi™ a 3D imager under development that returns a depth map from 2D video.