Can AI robots replace human astronauts?

On Christmas Eve, an autonomous spacecraft flew past the Sun, closer than any human object before it. Flying through the atmosphere, NASA’s Parker Solar Probe was on a mission to discover more about the Sun, including how it affects space weather on Earth.

This was a pivotal moment for humanity – but without any direct human involvement, as the spacecraft carried out its pre-programmed tasks on its own as it flew past the sun, without any communication with Earth.

Robotic probes have been sent across the solar system for the past six decades, reaching destinations impossible for humans. During the ten-day flightthe Parker solar probe experienced temperatures of 1000C.

But the success of these autonomous spacecraft—along with the rise of new advanced artificial intelligence—raises the question of what role humans might play in future space exploration.

Some scientists question whether human astronauts will even be needed.

“Robots are developing rapidly and the case for sending humans is getting weaker,” says Lord Martin Rees, Britain’s Astronomer Royal. “I don’t think taxpayer money should be used to send people into space.”

It also indicates a risk to humans.

“The only case for sending people [there] is like an adventure, an experience for rich people, and it should be financed privately,” he claims.

Andrew Coates, a physicist at the University of London, agrees. “For serious space exploration, I prefer robotics,” he says. “[They] go much further and do more things.”

They are also cheaper than humans, he claims. “As artificial intelligence advances, robots can get smarter and smarter.”

But what does this mean for future generations of aspiring astronauts – and surely there are certain functions that humans can perform in space that robots, however advanced, could never do?

Robotic spacecraft have visited all the planets in the solar system, as well as many asteroids and comets, but humans have only gone to two destinations: Earth’s orbit and the Moon.

In total, about 700 people have been in space, the earliest in 1961, when Yuri Gagarin from the then Soviet Union became the first space explorer. Most of them were in orbit (circling the Earth) or suborbit (short vertical jumps into space lasting a few minutes, on vehicles like the New Shepard rocket of the American company Blue Origin).

“Prestige will always be the reason why we have people in space,” says Dr. Kelly Weinersmith, a biologist at Rice University in Texas and co-author of A City on Mars. “It seems to have been arranged as a great way to show that your political system is efficient and your people are brilliant.”

But aside from an innate desire to explore or a sense of prestige, humans also conduct research and experiments in Earth orbit, such as the International Space Station, and use them to advance science.

Robots can contribute to these scientific investigations, with the ability to travel to locations inhospitable to humans, where they can use instruments to study and probe the atmosphere and surfaces.

“Humans are more versatile and we do things faster than robots, but it’s really difficult and expensive to keep us alive in space,” says Dr. Weinersmith.

In her 2024 Booker Prize-winning novel Orbital, author Samantha Harvey puts it more lyrically: “A robot has no need for hydration, nutrients, excretion, sleep… It wants and asks for nothing.”

But there are also downsides. Many robots are slow and methodical – on Mars, for example, rovers (motorized remote control vehicles) run at a speed of barely 0.1 miles per hour.

“AI can beat humans at chess, but does that mean it will be able to beat humans at environmental exploration?” asks dr. Ian Crawford, a planetary scientist from the University of London. “I just don’t think we know.”

He believes, however, that artificial intelligence algorithms could allow rovers to be “more efficient.”

Technology can play a role in complementing human space travel by freeing astronauts from certain tasks to allow them to focus on more important research.

“[AI could be used to] to automate tedious tasks,” explains Dr. Kiri Wagstaff, a computer and planetary scientist from the US who previously worked at NASA’s Jet Propulsion Laboratory in California. “On the surface of a planet, people get tired and lose focus, but machines don’t. “

The challenge is that enormous amounts of power are required to run systems like large-scale language models (LLMs), which can understand and generate human language by processing vast amounts of textual data. “We are not at the point where we could start an LLM on a Mars rover,” says Dr. Wagstaff.

“The rover’s processors run about one-tenth [of the speed] that your smartphone has” – meaning they can’t cope with the intense demands of running an LLM.

Complex humanoid machines with robotic arms and limbs are another form of technology that could take over basic tasks and functions in space, particularly as they more closely mimic the physical capabilities of humans.

NASA’s Valkyrie robot was built by the Johnson Space Center to compete in the 2013 Robotics Challenge. Weighing in at 300lb and standing 6ft 2in tall, he looks like a Star Wars Stormtrooper, but is one of a growing number of humanoid machines with superhuman abilities.

Long before Valkyrie was created, Nasa Robonaut was the first humanoid robot designed for use in space, taking over tasks normally performed by humans.

Its specially designed hands meant it could use the same tools as astronauts and perform complex, delicate tasks such as grasping objects or flicking switches that were too demanding for other robotic systems.

A later model Robonaut flew to the International Space Station aboard the Discovery spacecraft in 2011, where it assisted in maintenance and assembly.

“If we need to change a component or clean a solar panel, we could do it robotically,” says Dr. Shaun Azimi, head of the dexterous robotics team at NASA’s Johnson Space Center in Texas. “We see robots as a way to secure those habitats when people aren’t around.”

He argues that robots could be useful, not to replace human researchers, but to work alongside them.

Some robots are already working on other planets without humans, sometimes even making their own decisions. NASA’s Curiosity rover is, for example exploring a region called Gale Crater on Mars and autonomously performs some of its science without human input.

“You can direct the rover to photograph the scene, look for rocks that might match the science priorities of the mission, and then autonomously fire your laser at that target,” says Dr Wagstaff.

“He can get a reading of a particular stone and send it back to Earth while people are still asleep.”

But the capabilities of rovers like Curiosity are limited by their slow pace. And there’s something else I can’t compete with. That is, humans have the added bonus of inspiring people on Earth in a way that machines cannot.

“Inspiration is something that is intangible,” claims prof. Coates.

Leroy Chiao, a retired NASA astronaut who went on three space flights in the 1990s and 2000s on NASA’s Space Shuttle and the International Space Station, agrees. “People connect when people do something.

“The public is excited about robotic missions. But I would expect the first man on Mars to be even bigger than the first moon landing.”

Humans have not traveled beyond Earth’s orbit since December 1972, when the last Apollo mission visited the moon. NASA hopes to return humans there this decade with its Artemis program.

The next manned mission will see four astronauts fly around the moon in 2026. The next mission, scheduled for 2027, will see NASA astronauts land on the lunar surface.

The Chinese space agency, meanwhile, also wants to send astronauts to the moon.

Elsewhere, Elon Musk, CEO of the American company SpaceX, has his own plans for space. He said his long-term plan is to create a colony on Mars, where humans could land.

His idea is to use the Starship, a large new vehicle his company is developing, to transport up to 100 people at a time, with the aim of a million people on Mars in 20 years.

“Musk claims that we need to move to Mars because it could be a backup for humanity if something catastrophic happens on Earth,” explains Dr. Weinersmith. “If you accept that argument, then sending people into space is necessary.”

However, there are huge unknowns about life on Mars, including a myriad of technical challenges that she says remain unsolved.

“Maybe babies can’t develop in that kind of environment,” she says. “There [are] ethical issues [like this] to which we have no answers.

“I think we should slow down.”

However, Lord Rees has his own vision where human and robotic research could merge to the point where humans themselves are partly machines to deal with extreme environments. “I can imagine they’ll use all the genetic modification techniques, cyborg augmentations and so on, to deal with very hostile environments,” he says.

“Maybe we’ll have a new species that will happily live on Mars.”

Until then, however, humans will likely continue their small steps into the cosmos, a path trodden long ago by robotic explorers before them.

Main image credit: NASA

BBC InDepth is the new home on the website and app for the best analysis and expertise from our top journalists. Under a distinctive new brand, we’ll bring you fresh, thought-provoking perspectives and in-depth reporting on the biggest issues to help you make sense of a complex world. We’ll also feature thought-provoking content from BBC Sounds and iPlayer. We’re starting small but thinking big and we want to know what you think – you can send us your feedback by clicking the button below.