Opteran’s general-purpose neuromorphic software could be heading to Mars in the coming years. The company today announced that Airbus Defence and Space, with support from European Space Agency and the UK Space Agency, will test the Opteran Mind in Airbus space rovers.
London-based Opteran said that nature offers a more efficient, robust approach to autonomy in space robotics. The company said its product will enable new mission capabilities for future Mars missions and other space exploration projects.
Based on more than a decade of research into animal and insect vision, navigation, and decision-making, Opteran is conducting tests with Airbus at its Mars Yard to give rovers depth perception in the toughest off-world environments.
“Mars is possibly one of the hardest places to do autonomy,” David Rajan, co-founder and CEO of Opteran, told The Robot Report. “Doing autonomy, in general, is typical on this planet, but we’re trying to do it on a planet where it takes 30 minutes or so, from Mars, to get a message.”
“It’s not like the moon, which is very, very close,” he added. “I think people don’t realize quite how far away Mars is in comparison to the moon. So, working in this area with a firm trying to build a machine to be autonomous in a place that is so far away and so hostile was just instantly interesting.”
Opteran said it has reverse-engineered natural brain algorithms into software that enables autonomous machines to efficiently move through challenging environments without the need for extensive data or training. Founded in 2020, the company earned a 2024 RBR50 Robotics Innovation Award for this work.
Robustness is paramount for space robotics
“Part of the [challenge] is the cost and complexity of building a system that you’re going to send to Mars. Just getting it there is so expensive,” noted Rajan. “So, if it breaks, you can’t just send another one out. You certainly can’t send another person out there to fix it. So the systems have got to be robust.”
Many of today’s off-world robots are cumbersome — taking minutes to compute a map of their surroundings from multiple cameras before every movement.
“Today, when you’re building a panoramic depth map, it can take minutes to process the data from the sensor and decide what you’re going to do,” Rajan explained. “So the machine, the rover, is essentially stopping to process the data, which takes a few minutes, and then it can work out what to do and then move again.”
These rovers need to repeat this process every few feet, or even inches, depending on the area, said Rajan. This means the rover takes slow, baby steps across the planet.
Opteran claimed that its visual and perception systems offer Mars rovers the ability to understand their surroundings in milliseconds, in challenging conditions, without adding to the robot’s critical power consumption.
“If you’re going to shoot it on a rocket, then the weight of the machine really matters,” Rajan said. “And if it’s going to be running remotely off of batteries, probably solar charging, then power consumption really, really matters.”
“So it’s a low-weight algorithm that’s low power, and an algorithm that can process and build a depth map at 90 frames per second. So it’s kind of instant,” Rajan continued. “That completely transforms the pace and what you can do. Essentially, we take minutes of processing to nothing.”
Opteran discusses future plans for Mars exploration
Opteran said that the successful application of its Opteran Mind technology to real-world space exploration will significantly extend navigation capabilities in extreme off-world terrain. Ultimately, this provides rovers with continuous navigation while being able to drive further and faster.
The project’s near-term focus is on depth estimation for obstacle detection, and the mid-term focus on infrastructure-free visual navigation.
“What we do on the ground with customers is to enable autonomous machines to move independently without infrastructure. This means being able to judge not only depth, but also to locate, build maps, and navigate,” Rajan said.
“This is not part of this project. I should be clear about that, but our aspiration would go beyond depth perception to look at localization and mapping,” he continued.
Rajan said Opteran is interested in someday working with smaller, lighter, and cheaper rovers that could be sent to Mars as a fleet. These more disposable robots could tackle the mapping of difficult terrains.
Once the results of the initial testing have been presented to the European Space Agency (ESA) the goal would be to move to the next stage of grant funding, which would start to focus on deployment and commercialization.
“It’s a team effort, and every tiny aspect has to be considered, and it has to be the best it can be,” Rajan said. “It has to be optimized to oblivion, because it’s going to face some of the harshest conditions, not just on the planet but also in the solar system.”
This project is funded by ESA’s General Support Technology Programme (GSTP) through the UK Space Agency, which takes leading-edge technologies that are not ready to be sent into space and then develops them to be used in future missions.