A new method could enable multi-robot teams to explore other planets autonomously and reliably

This article was reviewed according to Science

fact checked

preprint

trusted source

proofread


The rocker bogie runt rover was chosen because it has a small form factor and uses off-the-shelf components. Credit: Sarah Swinton.

× close to


The rocker bogie runt rover was chosen because it has a small form factor and uses off-the-shelf components. Credit: Sarah Swinton.

Although roboticists have developed increasingly sophisticated systems in recent decades, it is often a challenge to ensure that these systems can function autonomously and without accidents in practice. This is especially difficult when these robots are designed to be deployed in complex environments, including space and other planets.

Researchers from the University of Glasgow recently developed a new methodology that allows teams of multiple rovers to explore other planets autonomously and reliably. This method was introduced in an article pre-published at arXivcontains data taken from various sources, including image data, maps and information collected by sensors, to plan efficient routes for different robots in a team.

“Using a team of planetary exploration rovers to explore the surface of Mars, rather than a single rover, could greatly expand the scientific capabilities of a mission,” Sarah Swinton, first author of the paper, told Tech Xplore. “All planetary exploration rovers must maintain some level of autonomy, because communication latencies between Earth and Mars make it extremely difficult and time-consuming for humans to perform propulsion actions. Deploying a team of rovers places a further emphasis on autonomy, as the difficulty of coordinating their behavior increases for human operators.”

The primary goal of Swinton and her collaborators’ recent study was to effectively address a long-standing research problem in robotics: how to effectively tackle autonomous multi-robot planetary exploration missions. To do this, the team has developed a multi-rover mission planner that will allow a team of several rovers, small robots designed for space exploration, to autonomously, safely and efficiently explore an area of ​​the Martian surface.

“The method we propose allows a robotic team to autonomously explore the surface of Mars through two key phases: map generation and mission planning,” Swinton explains. “First, a map of the area is created using data from the Mars Reconnaissance Orbiter. We specifically used data from Jezero Crater, where NASA’s Perseverance rover is currently operating.”

After the team’s planner creates a map of the environment the rovers will explore on Mars, it analyzes it and divides it into different regions, marking areas of terrain that the rovers can safely traverse. The planner then overlays a probability distribution map, which highlights the likelihood that rovers will encounter sites of scientific interest at specific locations in the environment they are exploring.

“These points could represent rocks that we want the rovers to sample from,” Swinton said. “Once this map is created, our mission planner searches the environment to identify an efficient route that increases the likelihood of finding the points of interest. Then, a coordinated set of safe paths for each member of the rover team is identified.”

The multi-rover mission planner developed by Swinton and her colleagues has several advantages over previously developed approaches. In addition to defining terrain that the rovers can safely travel in and planning routes for their autonomous operation, the planner also provides information about where locations might be of scientific interest.

“Our rover team is able to safely and efficiently search an entire mission site spanning 75,000 feet2 in a relatively short time,” Swinton said. “It is also worth noting that each rover will complete an autonomous driving distance comparable to the current record for ‘longest distance driven without human review’ by a planetary exploration rover. Our work also showed that search efficiency was improved by deploying a rover team instead of a single rover.”

Swinton and her colleagues evaluated their mapping and planning approach in a series of tests and simulations, conducted using a series of randomly generated probability distribution maps. Their results were promising, suggesting that their method could enable a team of five rovers to autonomously explore an area of ​​22,500 meters.2 on Mars in about 40 minutes.

Although the planner has so far been applied to Mars exploration, it could be applied to missions other than planetary exploration. For example, it could also help coordinate the efforts of multiple ground robots during search and rescue operations, simply by using a map of the area in question and a probability distribution map that marks locations where the robots are most likely to encounter people in need of rescue . or who need help.

In their next research, Swinton and her colleagues plan to further develop and test their methodology, while also working on other computational tools to support the autonomous operation of multiple robots. These tools also include methods to improve the fault tolerance of multi-robot teams.

“The consequences of errors and failures are a serious concern in planetary exploration missions,” Swinton added. “For a team of planetary exploration robots to be considered reliable, the robots must be able to diagnose faults in themselves and/or their teammates. Only once the errors have been diagnosed can remedial actions be taken to limit the impact the error has on the mission. outcomes.”

More information:
Sarah Swinton et al, A new methodology for autonomous planetary exploration using multi-robot teams, arXiv (2024). DOI: 10.48550/arxiv.2405.12790

Magazine information:
arXiv

Leave a Comment