Autonomous Race Drones to Revolutionize Spacecraft Steering

In a groundbreaking study published in Science Robotics, researchers from the European Space Agency (ESA) and the Technical University of Delft (TU Delft) have explored the use of autonomous race drones to revolutionize spacecraft steering. The study, titled “Optimality principles in spacecraft neural guidance and control,” has shown that these tiny drones can be used to test neuronal network-based guidance and control systems for spacecraft, paving the way for more efficient and robust missions.

Background and Challenges

Currently, spacecraft steering is done using pre-planned trajectories and carefully calculated maneuvers. However, this approach has limitations, as it does not take into account unforeseen events or changes in the mission plan. Autonomous systems that can adapt to changing conditions and optimize their performance in real-time are needed to overcome these challenges.

Neuronal Networks for Optimal Steering

The researchers have proposed using neuronal networks, specifically called Guidance & Control Networks (G&CNets), to control spacecraft. These networks can learn from experience and adapt to changing conditions, making them ideal for autonomous systems. The G&CNet should be able to directly control the actuators of the spacecraft, such as thrusters and propellers, and navigate the spacecraft through complex maneuvers.

Real-World Testing with Race Drones

To test their theory, the researchers have used race drones as a proof-of-concept. These tiny drones are equipped with propellers and can fly at high speeds. The researchers have programmed the drones to follow a predefined path, and then tested the G&CNet’s ability to adapt to changing conditions and optimize their performance.

The Results

The results were promising, with the G&CNet-controlled drones able to navigate through complex maneuvers and adapt to changing conditions in real-time. The researchers were able to validate their theory that neuronal networks can be used for optimal steering of spacecraft.

Future Plans

The next step for the researchers is to build a demonstrator for a world space mission. They plan to use the race drones as a starting point and integrate the G&CNet into the spacecraft’s control system. The demonstrator will be tested in a real-world environment, paving the way for more efficient and robust space missions.

Conclusion

The use of autonomous race drones to test neuronal network-based guidance and control systems for spacecraft has shown promising results. These tiny drones have the potential to revolutionize spacecraft steering, enabling more efficient and robust missions. The future of space exploration looks bright with this technology on the horizon.