Scientists create the first living robot with the ability to reproduce

Scientists have succeeded in building the first living robots capable of replicating and reproducing, raising hopes for fighting diseases such as the next possible pandemic and cancer.

Xenobots, living robots designed with the help of artificial intelligence (AI), represent an entirely new form of biological self-replication that holds promise for regenerative medicine.

To survive, life must reproduce, and over billions of years, living things have evolved many ways to replicate; From germinating plants to sexual animals and invasive viruses. Now, however, scientists have discovered a completely new form of biological reproduction and have used it to create the first living and self-replicating robots.

The same research team that in 2020 made the first living robots, i.e. xenobots, by collecting frog cells, has recently made a new discovery about their replication. The team found that these hand-assembled, computer-designed creatures could swim around in their tiny container, find individual cells, collect hundreds of them, and “baby” xenobots inside their mouths. Make your own Pacman-like.

After a few days, these babies become complete new xenobots, looking and moving exactly like the originals. These new living robots can then move out of their confines, find the cells, and make copies of themselves again. A process that repeats itself over and over again.

The results of recent research have been published in the Proceedings of the National Academy of Sciences. “With the right design of the robot, they will reproduce spontaneously,” said Joshua Bongard, a computer scientist and robotics expert at the University of Vermont who led the new research.

into the unknown

In the “African Clawed Frog” or “Xenopus Laevis” these embryonic stem cells turn into skin. “They sit on the outside of a tadpole, keeping pathogens out and redistributing mucus,” said Michael Levin, professor of biology and director of the Allen Discovery Center at Tufts University. “But we’re putting them in a whole new context and giving them a chance to rethink their multicellularity.”

And what they imagine is something very different from the performance on the skin. “Douglas Blackiston,” one of the researchers in this field who collected the parents of xenobots and developed the biological part in the recent study, said: “People thought for a long time that we knew all the ways that life can reproduce or To repeat, we have tried it. But this is something that has not been seen before.”

Levin also noted: “This is a very deep issue. “These cells have the genome of a frog, but once they break free from becoming a frog, they use their collective intelligence to do something amazing.”

In previous experiments, scientists were surprised that xenobots could be designed to accomplish simple tasks. Now they are amazed that these biological objects, or in other words a set of cells designed by a computer, will reproduce automatically.

“We have the complete, unaltered genome of the frog, but there’s no indication that these cells can work together to do this new thing, which is to collect and then compress individual cells to make a copy,” Levin said.

“These are frog cells, but they replicate in a very different way than frogs do,” said lead author of the new study, Sam Kriegman, who received his Ph.D. from the University of Vermont. “No animal or plant known to science reproduces in this way.”

The zenobot parent, which is made of about 3000 cells, forms a sphere by itself. “These can produce babies, but then the system usually collapses,” Kriegman said. “In fact, it is very difficult to force a living robot system to continue replicating.”

But with an artificial intelligence program running on the “Deep Green” supercomputer cluster at the University of Vermont’s (UVM) Advanced Computing Core, the development of an evolutionary algorithm was able to simulate billions of body shapes such as triangles, squares, pyramids, and starfish. Experiment and find shapes that allow the cell to be more efficient in motion-based kinematic replication.

The baby’s parent cell makes a copy of itself by collecting the stem cells into a ball.
Credit: Douglas Blackiston and Sam Kriegman

“We asked UVM’s supercomputer to figure out how to adjust the parent’s shape, and after months of trying, the AI ​​came up with strange designs, including one that looked like Pac-Man,” Kriegman added. This is very counterintuitive. Although the design looks very simple, it is not something that a human engineer could handle. Why a small mouth? Why not five? We sent the results to Douglas Blackstone, and based on them in the lab, he built these Pacman-shaped xenobot parents. Then these parents produced children who produced grandchildren and further descendants were produced.” In other words, proper design significantly increased the number of generations.

In fact, kinetic replication is well known at the level of molecules, but it has never been seen before at the scale of cells or whole organisms, and now it is possible with living robots.

“We discovered that this previously unknown space exists in living organisms or systems, and it’s a vast space,” said Bongaard, a professor in UVM’s School of Engineering and Mathematical Sciences. So how can we explore it? We found xenobots that walk. We found zenobots that swim. And now in this study we have found xenobots that repeat kinetically. What else is there to see in this realm?”

As the scientists wrote in a recent study, “Life reveals amazing behaviors just beneath the surface, waiting to be discovered.”

The method of replicating the Zenobot living robot
redit: Doug Blackiston and Sam Kriegman

Response to risk

Some people may find this finding exciting. Others may react with concern or even horror to the concept of self-replicating biotechnology. But for the team of scientists, the next goal is to understand the subject more deeply.

“We’re working to understand this feature: replication,” Bongaard said of the development of living robots. “The world and technologies are changing rapidly and it is important for the whole society to study and find out how this phenomenon works.”

These millimeter-sized living machines are completely lab-based, easily shut down, and have been reviewed by federal, state, and institutional ethics experts.

“These are not things that worry me,” said the researcher. What creates the risk is the next pandemic, the acceleration of ecosystem damage from pollution, the intensification of threats from climate change.”

“This is an ideal system in which to study self-replicating systems,” he added. “We have an ethical imperative to understand the circumstances under which we can control, direct, stop or develop experiments.”

The research team of this study from left: Josh Bongard, Michael Levine, Douglas Blackiston, Sam Kriegman
Credit: Tufts and ICDO

Bongard mentioned the corona epidemic and vaccine development. “The speed with which we can generate solutions is deeply important,” he added. If we can develop technologies by learning from xenobots, we can quickly tell the AI ​​that we need a biological tool that does x and y and suppresses z. This can be very beneficial. Which now takes a very long time.”

“The goal of this research team is to speed up how quickly people can go from identifying a problem to generating a solution,” he noted. Solutions like deploying living machines to pull microplastics out of waterways or making new drugs. “We need to create technological solutions that grow as much as the challenges we’re dealing with.”

And this research team sees hopes for progress in the field of regenerative medicine in their research. “If we figure out how to tell a set of cells to do what we want them to do, which is ultimately regenerative medicine, that could be a solution for traumatic injuries, birth defects, cancer and aging,” Levin said. All these various problems exist because we don’t know how to predict and control the groups of cells that are going to be made. But Xenobots are a new platform to teach us.”

Cover photo: The organism designed with artificial intelligence produces a stem cell in the form of a compressed ball.
Credit: Douglas Blackiston and Sam Kriegman

Source: SciTechDaily