The majority of current engineered approaches to calculating self-motion from visual information generally relies on object identification and/or correlation of image sections between successive frames. Such calculations are computationally expensive, require high-frame rate cameras for smooth motion detection and real-time digital implementation requires powerful processors. Alternatively, flies arrive at essentially the same outcome, estimation of self-motion, in a much smaller package using vastly less power. Despite potential advantages, and some notable successes, few visual neuromorphic analog VLSI devices have been employed. Due to various practical limitations most devices are promising in simulation but do not work as expected when put into real-world testing. This paper describes a hardware implementation, in aVLSI, of an adaptive model for motion detection. The chip integrates motion over a 1D circular array of local motion processors to give a single voltage output. The device is integrated with a circular array of 40 photodiodes to form a visual rotation sensor. Tested using natural images, this motion detecting chip provides an accurate estimation of rotational motion largely independent of the structure and contrast of the scene, using <1 mW of power and weighing only a few grams. In principle, such devices could be used as an element of a velocity-based servo to replace or augment inertial guidance systems in applications such as mUAVs, with wider applications as smart car sensors, intruder detectors and even visual aids for the blind.