MIT researchers have successfully developed advanced algorithms aimed at optimizing the trajectory planning and control of tailsitter, a fixed-wing drone design known for its vertical take-offs and landings. With origins traced back to a design by Nikolai Tesla in 1928, the tailsitter's performance enhancement becomes increasingly significant due to its efficiency over conventional quadcopter drones.
Unlike traditional aircraft, tailsitters can operate like a helicopter and an airplane, making them ideal for roles like parcel delivery or search-and-rescue missions. Their design allows for versatile maneuvers, transitioning between vertical and horizontal flight, and even sideways or inverted positions.
The new algorithms, spearheaded by MIT's Ezra Tal, are designed to improve the aircraft's ability to perform complex maneuvers in real-time. Previously, trajectory planning either oversimplified the dynamics or utilized two different models depending on the aircraft's mode (airplane or helicopter). MIT's approach merges these into one comprehensive model.
Central to the new system's efficiency is the use of differential flatness, a technical property that allows for rapid trajectory feasibility checks. Traditional methods required extensive calculations to determine if certain flight paths were possible, but with differential flatness, the team can quickly and efficiently confirm trajectory viability.
The MIT team showcased the potential of their algorithms through various tests, including a synchronized "airshow" of three tailsitters executing complex aerial maneuvers. This new development could be a breakthrough in drone technology, given the tailsitter's efficiency in forward flight and its past challenges tied to manual piloting.
With their current success in indoor environments, the team is now focusing on adapting their algorithms to handle external factors like wind for outdoor flight. The research, detailed in IEEE Transactions on Robotics, received backing from the U.S. Army Research Office.