Insects flying abilities based on optic flow (OF) are nice bio-inspired models for Micro Aerial Vehicles (MAVs) endowed with a limited computational power. Most OF sensing robots developed so far have used numerically complex algorithms requiring large computational power often carried out offline. The present study shows the performances of our bio-inspired Visual Motion Sensor (VMS) based on a 3x4 matrix of auto-adaptive aVLSI photoreceptors pertaining to a custommade bio-inspired chip called APIS (Adaptive Pixels for Insect-based Sensors). To achieve such processing with the limited computational power of a tiny microcontroller (μC), the μC-based implementation of the "time of travel" scheme requiring at least a 1kHz sampling rate was modified by linearly interpolating the photoreceptors signals to run the algorithm at a lower sampling rate. The accuracy of the measurements was assessed for various sampling rates in simulation and the best trade-off between computational load and accuracy determined at 200Hz was implemented onboard a tiny μC. By interpolating the photoreceptors signals and by fusing the output of several Local Motion Sensors (LMSs), we ended up with an accurate and frequently refreshed VMS measuring a visual angular speed and requiring more than 4 times less computational resources.