Robots to collect dust on the moon

Created by the University of Bristol's researchers at the Bristol Robotics Laboratory, this innovative tool holds promise for astronaut training before Lunar missions.

In collaboration with Thales Alenia Space in the UK, who has a vested interest in developing robotic systems for space exploration, the team explored a virtual rendition of regolith, commonly referred to as Moon dust.

With upcoming Lunar exploration missions on the horizon, Lunar regolith garners significant interest. It is believed that regolith could be a source of essential resources like oxygen, rocket fuel, or materials for construction, supporting sustained human activity on the Moon.

For the collection of regolith, remotely operated robots are considered a viable option due to their reduced risks and costs compared to manned space missions. Nevertheless, the considerable distance these robots operate over introduces significant delays, complicating their control.

With the simulation's proven accuracy in mimicking real-world behaviors, it offers a practical solution for operating robots on the Moon, eliminating delays and ensuring smoother, more efficient control.

Joe Louca, the study's lead author from Bristol’s School of Engineering Mathematics and Technology, likens the simulation to a highly realistic Moon-based video game. The goal is to ensure the virtual Moon dust behaves identically to the real thing, facilitating accurate robot control on the Lunar surface. This model is distinguished by its accuracy, scalability, and low computational demands, making it an asset for future Lunar missions.

Building on prior research, the team recognised that expert robot operators prefer training with increasing levels of risk and realism. Starting with simulations and progressing to physical mock-ups before handling the actual systems is essential for developing confidence and familiarity. Therefore, an accurate simulation model is vital for effective training.

Although highly detailed Moon dust models exist, their computational demands render them impractical for smooth robot control. To address this, researchers from the German Aerospace Centre (DLR) developed a more computationally efficient virtual regolith model that accounts for its properties and the Moon's lower gravity. However, this model's effectiveness is limited to smaller quantities of regolith.

The Bristol team aimed to enhance this model to manage larger quantities of regolith while maintaining its real-time operational capability, followed by experimental validation.

Joe Louca emphasised the project's focus on improving the operators' experience, making their tasks more manageable. The team refined DLR's initial model for greater scalability and conducted experiments to compare the virtual and real-world behavior of Moon dust.

Given the model's accuracy, scalability, and real-time operational capability, further exploration into its application for regolith collection is warranted. The team also contemplates extending this simulation approach to Martian soil, potentially benefiting future Mars exploration missions or training for the Mars Sample Return mission.