The new algorithm, which was developed by aerospace engineers at the Massachusetts Institute of Technology (MIT), was designed to find the fastest route around obstacles without crashing. It combines simulated and real flight data from a drone tackling an obstacle course in both physical and virtual space.
The faster drones fly, the more unstable they become, and at high speeds their aerodynamics can be too complicated to predict. Crashes are a common occurrence in autonomous drone racing.
If they can be pushed to be faster and nimbler, however, small autonomous aircraft could be used in ‘time-critical’ operations beyond the race course – searching for survivors after a natural disaster, for example.
Despite completing an obstacle course up to 20% quicker than drones using conventional algorithms, the new program did not always keep the drone ahead of competitors throughout the course. In some cases, it chose to slow down to handle a tricky curve, or to save its energy in order to speed up later and ultimately overtake its rival.
“At high speeds, there are intricate aerodynamics that are hard to simulate, so we use experiments in the real world to fill in those black holes to find, for instance, that it might be better to slow down first to be faster later,” said graduate student Ezra Tal. “It’s this holistic approach we use to see how we can make a trajectory overall as fast as possible.”
Aerodynamics such as drag do not generally come into play at low speeds, the researchers said, so they can be left out of modelling of a drone’s behaviour. At high speeds, however, such effects are far more pronounced and they make handling much harder to predict.
High speeds also bring other challenges. “When you’re flying fast, it’s hard to estimate where you are,” said graduate student Gilhyun Ryou. “There could be delays in sending a signal to a motor, or a sudden voltage drop which could cause other dynamics problems. These effects can’t be modelled with traditional planning approaches.”
To minimise the number of physical test flights – and crashes – needed to identify fast and safe flight paths, the team combined simulation and experiments. They started with a physics-based flight planning model, which they developed to first simulate how a drone is likely to behave while flying through a virtual obstacle course. They simulated thousands of racing scenarios, each with a different flight path and ‘speed pattern’. They then charted whether each scenario was feasible (safe), or infeasible (resulting in a crash). From this chart, they could quickly find a handful of the most racing trajectories to try out in the lab.
“We can do this low-fidelity simulation cheaply and quickly, to see interesting trajectories that could be both fast and feasible. Then we fly these trajectories in experiments to see which are actually feasible in the real world,” said Tal. “Ultimately we converge to the optimal trajectory that gives us the lowest feasible time.”
To demonstrate the new approach, the researchers simulated a drone flying through a simple course with five large, square-shaped obstacles arranged in a staggered configuration. They set up this same configuration in a physical training space, and programmed a drone to fly through the course at speeds and trajectories that they previously picked out from their simulations. They also ran the same course with a drone trained on a more conventional algorithm that does not incorporate experiments into its planning.
Overall, the drone trained on the new algorithm won every race, completing the course in a shorter time than the conventionally trained drone.
The researchers plan to fly more experiments, at faster speeds and with more complex environments, to further improve their algorithm. They are also considering incorporating flight data from human pilots who race drones remotely, and whose decisions and manoeuvres might help identify even faster flight plans.
“These kinds of algorithms are a very valuable step toward enabling future drones that can navigate complex environments very fast,” said associate professor Sertac Karaman. “We are really hoping to push the limits in a way that they can travel as fast as their physical limits will allow.”
The research was published in the International Journal of Robotics Research.
Want the best engineering stories delivered straight to your inbox? The Professional Engineering newsletter gives you vital updates on the most cutting-edge engineering and exciting new job opportunities. To sign up, click here.
Content published by Professional Engineering does not necessarily represent the views of the Institution of Mechanical Engineers.