Autonomous control algorithms, developed by researchers from the Institute of Aerospace Studies at the University of Toronto (You should use it), could make cargo transportation on the moon safer and more efficient for astronauts in the future.
Within the team, which is headed by MDA Space, professor Tim Barfoot and graduate student Alex Kravchiv are working on the technology, which will help Canada's lunar utility vehicle navigate between offloading points on future missions. This solves one of the key logistical problems after landing astronauts on the surface of the moon.
During lunar missions it is assumed, that the landing site of the spacecraft and the habitation module will be located at a distance of approximately five kilometers from each other. The landing site must be level, to ensure the shuttles arrive safely, while the base should be located behind natural shelters, which protect against radiation. This creates a logistical challenge: it is necessary to organize regular transportation of all equipment and cargo between the two points.
Unlike previous planetary missions, where rovers explored unknown terrain in different directions, the lunar utility vehicle will make multiple trips between fixed locations, delivering equipment and materials. This is the first case, when the space rover must repeat the same route, therefore, Barfoot's "learn and repeat" visual navigation system (teach-and-repeat) perfect for this type of mission.
This approach involves, that the rover first "learns" the route under the control of the operator, after which it can automatically play it an unlimited number of times. By automating part of the mission, astronauts will be able to save time and effort, limit the impact of lunar conditions on yourself and increase the productivity of the entire expedition.
As part of his research, Kravchev is adapting autonomous driving technology for integration with the Canadian Space Agency's Lunar Exploration Light Rover test rover. (LELR). In December 2024 In 2018, the UTIAS team, together with MDA Space and the University of Sherbrooke's BRP Center for Advanced Technologies, tested the autonomous system at the agency's Montreal test site, which simulates the terrain of Mars. During the tests, the researchers were able to evaluate the operation of the hardware and software in the conditions, close to menstruation.
While adapting the system to LELR, the team faced a number of challenges: simulation of lunar conditions created a delay of five seconds between sending a command and receiving a return signal, which made standard joystick control impossible. This prompted researchers to develop a new semi-autonomous learning method, which uses short route segments — an approach, which was not used before.
After successful trials in July 2025 In 2018, the team received a contract from the Canadian Space Agency for preliminary technical research of a lunar utility vehicle within the framework of the Lunar Surface Exploration Initiative program — Canada's contribution to the NASA Artemis program, aimed at creating a permanent human presence on the Moon.
Currently, researchers are preparing the vehicle for full-fledged missions, focusing on increasing system reliability and stability during long-term operations. They point out, that field testing provided valuable experience not only in the field of autonomous driving, but also in creating a user-friendly interface, which should work stably even in a complex environment. The obtained conclusions will form the basis of the next stage of development.
Source: https://phys.org