Planning system ensures a robot’s flight path will remain collision-free

A new autonomous navigation system, named SANDO (Safe AutoNomous trajectory planning for Dynamic unKnOwn environments), has been developed by MIT researchers to ensure a UAV’s flight path remains collision-free in unfamiliar and dynamic environments. This groundbreaking method allows UAVs to fly deep into a wildfire or other hazardous situations, gathering critical information for rescuers while avoiding sudden flare-ups, falling debris, and unpredictable obstacles.

Unlike popular navigation systems that can only offer safety guarantees for static or known environments, SANDO provides a mathematically proven guarantee of safety even when the environment is completely unknown and obstacles are moving. The system starts by mapping out a safety corridor through the environment, which is a series of connected regions of 3D space the robot can travel through without colliding with any obstacles.

A unique feature of SANDO is its consideration of the possible future movements of dynamic obstacles. By employing a module that detects, groups, and monitors these obstacles, the system estimates where they will move next and puts a sphere around each one, capturing the farthest distance they could travel in all directions. SANDO then builds the safety corridor around these spheres to ensure the UAV won't collide with moving objects.

Once a collision-free safety corridor is established, SANDO optimizes the trajectory within that corridor to find the fastest path to the goal. It adjusts the safety corridor and reformulates the trajectory as the robot travels, ensuring the path remains collision-free throughout the journey. SANDO's fast reactions and efficient optimization make it an ideal choice for high-stakes applications like search and rescue missions into collapsed buildings, mine explorations through hidden tunnels, or package delivery across a crowded neighborhood.