In the wake of a devastating earthquake, unmanned aerial vehicles (UAVs) could be used to navigate through collapsed buildings to survey the area, providing vital information to rescuers to locate survivors quickly. However, this poses a significant challenge for autonomous robots as they need to swiftly adjust their path to avoid sudden obstacles while maintaining their course.
Researchers from MIT and the University of Pennsylvania have developed a new trajectory-planning system that addresses both of these challenges simultaneously. This system allows a UAV to respond to obstacles in milliseconds while following a smooth flight path that minimizes travel time.
Their innovative technique uses a new mathematical approach to ensure the robot reaches its destination safely along a feasible path. It is also less computationally intensive compared to other methods, generating smoother trajectories faster than current state-of-the-art solutions.
The trajectory planner, named MIGHTY, is an open-source system that does not require expensive proprietary software packages, making it more accessible for deployment in various real-world scenarios. Apart from search-and-rescue operations, MIGHTY could also be beneficial for applications like last-mile delivery in urban spaces and industrial inspections of complex structures.
Overcoming trade-offs in existing systems, MIGHTY aims to democratize high-performance trajectory planning by achieving comparable or better results using open-source tools. This eliminates cost barriers and allows researchers, students, and companies worldwide to utilize the system freely.
The research team, led by Kota Kondo from MIT, developed MIGHTY with the goal of creating autonomous robots capable of maneuvering in hazardous environments without risking human lives. The system optimizes travel time and flight path together using a Hermite spline mathematical technique to produce a smooth and controllable trajectory.
By integrating everything into one piece of software, MIGHTY eliminates the need for communication with other software components, ensuring efficient and real-time obstacle avoidance while minimizing travel time. In simulated experiments, MIGHTY outperformed state-of-the-art methods by reaching its destination faster while using less computation time. Real robot tests confirmed MIGHTY’s ability to achieve high speeds while safely navigating around obstacles.
Looking ahead, the researchers aim to expand MIGHTY’s capabilities to control multiple robots simultaneously and conduct further flight experiments in challenging environments. They plan to enhance the open-source system based on feedback from users to continue improving its performance.
This research received partial funding from the United States Army Research Laboratory and the Defense Science and Technology Agency in Singapore.
