The success of autonomous driving is limited by the sensors, typically optically based, that map what is seen in front of the car. Both distance, direction, and the calculated relative velocities are critical to guiding a car to avoid other moving and stationary objects. So this paper describes an improved approach: http://web.media.mit.edu/~achoo/beat/Kadambi_beat_lowres.pdf Radar works by timing a reflected signal but the problem is sensor time resolution, they are too slow to achieve more than centimeter resolution. This paper outlines how to 'heterodyne' the transmitted and reflected light to achieve millimeter resolution. Better still, it appears to mitigate the fog problem. There is a big gap between lab and product but the math and physics laid out here is the roadmap to improved, autonomous driving sensors (and others.) Ones that may not require mechanical sweeping of the field. Bob Wilson
