BRAIN-POWERED SPIDER BOTS
Why are mini-brains being used to control robots?
Neuroscientist Alysson Muotri is putting organoids inside four-legged, spider-like robots. The main reason for such an unusual mind-machine interface is that it enables the ‘brains’ to explore the world around them – via an artificial body. “They’re similar to a newborn baby who’s trying to get used to touching things and experiencing different sensations,” Muotri explains.
Electrical activity rises as an embryonic brain develops but reaches a plateau before birth, needing external input from the senses for its neural circuits to mature. In a spider bot, electrodes capture activity from the organoid’s cells and split the electrical output between the robot’s limbs, which must then synchronise. “In a very simplistic way, that’s exactly what the human brain does when a baby is trying to crawl or walk,” says Muotri. “So we’re teaching the organoid how to make this spider robot walk, by coordinating movements of the legs.”
It’s already learning. The ‘brain’ can currently tell its body to move forward; the next step is teaching it to go backward when it encounters an obstacle, which requires providing feedback for when a movement is wrong. That brain training is being achieved in two ways. The first method is a form of ‘mind control’ called optogenetics, in which cells are genetically engineered to to be sensitive to light so that their behaviour can be controlled. The second method involves supplying cells with a dose of the reward hormone, dopamine. Both methods cause the organoid’s neural networks to rearrange, allowing it to adapt.
Once organoids are able to learn, they can battle in robot wars – obviously not a fight, but a competition of speed. An organoid made from Neanderthalised cells with a NOVA1 gene variant has defects in synaptic connections between its cells, affecting its neural networks. Would that make it slower to adapt compared to a human organoid? While speculative, the results could offer clues on whether our human ancestors were faster at thinking, influencing their behaviour.
The spider bot, with its multiple legs, is allowing Muotri’s lab to observe how a single organoid handles coordination. One follow-up experiment would be to see how several organoids deal with cooperation. Would they interact to accomplish a common goal? The design for cooperating robots is still in the planning stages – it might involve two or three human-like arms that must work together to lift an object, or robotic versions of social insects like ants and bees. “This might sound like science fiction,” says Muotri, “but it is actually happening, and they have a clear purpose.”
2 food to fuel an energy-hungry body in a cold climate. With modern humans, on the other hand, the trade-off meant that we had more processing power for social skills, contributing to our evolutionary success.
Ancient variants continue to affect humans today, as revealed by certain mental illness and neurological disorders. For example, ‘neurotypical’ people have one copy of a set of 25 genes located on chromosome 7, whereas a person with the condition ‘Dup 7’ has a duplication that creates two copies of each gene. This provides better visual and spatial skills but makes them less sociable – as in autism. At the other end of the spectrum, ‘Williams syndrome’ results from a deletion of those same 25 genes, leading to poor visual and spatial skills but hyper-sociability, which produces someone with a ‘gregarious brain’ who has a compulsion to strike up conversations with strangers.
So studying the impact of genetic mutations by comparing organoids, including those with ancient variants found in Neanderthals, can provide insights into conditions that affect modern humans. “This could provide information not only about what makes us who and what we are, but also possibly pave the way for treatment and diagnostic approaches,” explains Berman.
That’s actually the main reason why Muotri is growing mini-brains. His lab contains thousands of mutant organoids for various diseases. He’s already found that, like Neanderthalised structures, ‘autistic’ organoids have defects in their neural networks. “One misconception many people have is that by studying the evolution of the brain, there is no benefit to human health,” says Muotri, whose son suffers from autism. “By understanding how the social brain evolved, we might be able to create better therapeutics.”
“By understanding how the social brain evolved, we might be able to create better therapeutics”