Life must reproduce to persist. Plants, sexual animals, and invading viruses are just some of the ways in which organisms have evolved over billions of years.
The Wyss Institute for Biologically inspired Engineering at Harvard University and scientists at the University of Vermont have discovered a new form of biological reproduction that can be used to create self-replicating living robots.
The team that built the first living robot, called "Xenobots," assembled from frog cells, has discovered that these computer-designed and hand-assembled organisms can swim out into their tiny dish, find single cells, gather hundreds of them together, and assemble "baby" Xeno.
The new Xenobots can find and build their own copies of themselves. Again and again.
Joshua Bongard, a computer scientist and robotics expert at the University of Vermont, says that with the right design, they will spontaneously self-replicate.
The results of the new research were published in the journal.
Into the unknown.
The skin of a Xenopus laevis frog is made of embryonic cells. "They would be sitting on the outside of a tadpole, keeping out pathogens and redistributing mucus," says Michael Levin, a professor of biology and director of the Allen Discovery Center at Tufts University. We're putting them into a novel context. We are giving them a chance to rethink their multicellularity. The Wyss Institute has an Associate Faculty member named Levin.
They imagine something far different than skin. People have thought for a long time that we have figured out all the ways life can reproduce. Douglas Blackiston, the senior scientist at the Wyss Institute and the co-author of the new study, says that this is something that has never been observed before.
"This is profound," says Levin. "These cells have the genome of a frog, but they are free from becoming tadpoles and use their collective intelligence to do something amazing." The scientists were amazed at how easy it was to design Xenobots to accomplish simple tasks. They are shocked that these biological objects will spontaneously replicate. "We have the full, unaltered frog genome, but it gave no hint that these cells can work together on this new task," says Levin.
These are frog cells that are very different from the way they are made by the frog. The lead author of the new study, who completed his PhD in Bongard's lab at UVM, says that no animal or plant known to science replicates in this way.
The sphere formed by the Xenobot parent is made of 3,000 cells. The system usually dies out after these make children. It's difficult to get the system to keep reproducing. An evolutionary program was able to find the cells that were more effective at motion by testing billions of body shapes in simulation, thanks to an artificial intelligence program working on the Deep Green cluster at UVM's Vermont Advanced Computing Core.
After months of chugging away, the artificial intelligence at UVM came up with some strange designs, including one that looked like a video game. It's very hard to understand. It's not something a human engineer would come up with. Why is there one tiny mouth? Why not five? We sent the results to Doug, who built the parents. The parents built children, who built great-grandchildren, who built great-great-grandchildren. The right design extended the number of generations.
It has never been observed at the scale of whole cells or organisms.
Bongard says that there is a previously unknown space within organisms. How do we explore that space? We found a robot that walks. We found a creature that swims. In this study, we've found a way to replicate. What else is out there?
The scientists write in the study that "life harbors surprising behaviors just below the surface, waiting to be uncovered."
Complying with risk.
Some people may like this. Others may be concerned about the idea of a self-replicating technology. The goal is deeper understanding for the team of scientists.
We are trying to understand the property of replication. The world is changing fast. Bongard says that it's important for society as a whole to understand how this works. These millimeter-sized living machines are not what keep me awake at night. UVM's Bongard says that the risk is the next pandemic, the damage from pollution, and the threat from climate change. This system is ideal for studying self-replicating systems. We have a moral imperative to understand the conditions under which we can control it.
Bongard points to the COVID epidemic and the hunt for a vaccine. The speed at which we can come up with solutions is very important. If we can learn from XenoBots, where we can quickly tell the artificial intelligence that we need a biological tool that does X and Y, that could be very beneficial. That takes a long time today. Bongard says that the team wants to accelerate how quickly people can go from identifying a problem to generating solutions.
Bongard says that technological solutions need to grow at the same rate as the challenges they face.
The team sees promise in the research. "If we knew how to tell collections of cells to do what we wanted them to do, that's regenerative medicine -- that's the solution to traumatic injury, birth defects, cancer, and aging," says Levin. We don't know how to predict and control what groups of cells are going to build, that's one of the problems. There is a new platform for teaching.
The world's first self-replicating living robot is shown in this video.
The story was told
The University of Vermont has some materials. Joshua Brown wrote the original. Content can be edited for style and length.
