Nano-Drones: Enabling Indoor Collision Avoidance With a Miniaturized Multi-Zone Time of Flight Sensor

Unmanned aerial vehicles (UAVs) are active research topics, especially the nano and micro subclass. They are centimeter-size drones with minimal onboard computational capabilities. These light-weights platforms provide good agility and movement freedom in indoor environments(i.g. GPS-denied environments). However, it is still a significant challenge due to the limited onboard computational capabilities to enable autonomous navigation (or even primary obstacle avoidance abilities) using standard image sensors. Vision-based perception algorithms used routinely on standard-size drones prevent their use with state-of-the-art nano-UAVs. This work demonstrates the possibility of using a new multi-zone Time of Flight (ToF) sensor to enhance autonomous navigation with a significantly lower computational load rather than most common vision-based solutions. In particular, the novel integrated ToF sensor is characterized for the first time in literature in-field using an ad hoc light-weight PCB and the Crazyflie nano-UAV. Furthermore, an optimized approach to calculate collision probability from each ToF sensor frame has been proposed relying on empirical data. The main goal is to develop a new methodology capable of obstacle avoidance fusing only one 8*8 ToF on the front. The final system proved reliable (>95%) in-field obstacle avoidance capabilities when flying in indoor environments with dynamic obstacles.