The Hybrot, a small robot that moves about using the brain signals of a rat, is the first robotic device whose movements are controlled by a network of cultured neuron cells.
Steve Potter and his research team in the Laboratory for Neuroengineering at the Georgia Institute of Technology are studying the basics of learning, memory, and information processing using neural networks in vitro. Their goal is to create computing systems that perform like the human brain.
Potter, a professor at Georgia Tech and Emory University, presented his most recent findings last month during the Third International Conference on Substrate-Integrated Microelectrodes in Texas.
As the lead researcher on a $1.2 million grant from the National Institutes of Health, Potter is connecting laboratory cultures containing living neurons to computers in order to create a simulated animal, which he describes as a “neurally-controlled animat.”
“We call it the ‘Hybrot’ because it is a hybrid of living and robotic components,” he said. “We hope to learn how living neural networks may be applied to the artificial computing systems of tomorrow. We also hope that our findings may help cases in which learning, memory, and information processing go awry in humans.”
Inside Potter’s lab, a droplet containing a few thousand living neurons from rat cortex is placed on a special glass petri dish instrumented with an array of 60 micro-electrodes. The neurons are kept alive in an incubator for up to two years using a new sealed-dish culture system that Potter developed and patented. The neural activity recorded by the electrodes is transmitted to the robot, the Khepera, made by K-Team S.A, which serves as the body of the cultured networks. It moves under the command of neural activity that is being transmitted to it, and information from the robot’s sensors is sent back to the cultured net in the form of electrical stimuli.
Central to the experiments is Potter’s belief that over time, the team will be able to establish a living network system that learns like the human brain.
The team is able to make detailed observations of the neural signaling patterns, and document changes in the morphology and connectivity of the cells and networks by using high-speed cameras and voltage-sensitive dyes, in conjunction with 2-photon laser-scanning microscopy. The team is looking for evidence that the networks are growing and learning over time.
“Learning is often defined as a lasting change in behavior, resulting from experience,” Potter said. “In order for a cultured network to learn, it must be able to behave. By using multi-electrode arrays as a two-way interface to cultured mammalian cortical networks, we have given these networks an artificial body with which to behave.”