Scientists at Harvard University have created tiny living robots with nervous systems, offering a new way to study how nerve cells grow and organize inside unusual living bodies.
The study, led by Haleh Fotowat of the Wyss Institute for Biologically Inspired Engineering at Harvard University in the United States, was published in Advanced Science.
It describes a new kind of biological machine called a “neurobot.” These living robots are made from frog embryo cells and can move on their own through water.
Living robots with nervous systems take shape in Harvard Lab
Researchers built the neurobots from tissue taken from Xenopus laevis, a species of frog often used in laboratory research. Earlier work had already shown that frog skin tissue could form small, self-powered biological structures that move using tiny hair-like parts called cilia.
In the new study, researchers added neural precursor cells, which can develop into nerve cells, to those moving structures. The result was a living robot with a simple but real nervous system.
The implanted cells matured into neurons and sent out projections inside the body of the neurobot and toward its outer surface, the study found.
Scientists have created a “living robot” from cells capable of forming their own nervous system.
It can move, respond to stimuli, and be programmed; the era of programmable biology is approaching. pic.twitter.com/RfFjRdu8cC
— Tom Marvolo Riddle (@tom_riddle2025) April 6, 2026
Researchers also detected signs of synapse-like structures, which are the contact points that let nerve cells communicate. Calcium imaging, a method used to track cell activity, showed that the neurons were active.
Implanted neurons changed shape, structure, and motion
Those added nerve cells appeared to change how the bots looked and moved. Neurobots were often larger and less round than similar bots made without neural tissue.
They also showed more varied movement patterns. While standard biobots often moved in simpler paths, neurobots were more likely to follow more complex and changing routes.
Researchers also tested how the bots responded to a neuroactive drug known as PTZ. The drug affected neurobots and non-neural biobots in different ways. That difference suggested that the neurons were doing more than just growing inside the structures. They may also have been shaping behavior.
Gene activity points to deeper biological complexity
The study went further by examining gene activity. Neurobots showed stronger activity in genes linked to nervous system development, synapses, and neuron projections. Researchers also found higher expression of genes tied to visual perception.
That finding raises the possibility that these constructs may someday help scientists explore how simple living systems respond to light, though the study did not claim that as a confirmed behavior.
The researchers said the work offers a fully biological model for studying how nervous systems adapt when placed inside new kinds of bodies. The neurobots assemble themselves, move without outside power, and contain only living tissue.
The study does not present a ready-made technology for practical use. But it opens a new line of research into how neurons behave in unfamiliar forms of life-like machines.
For scientists, that could offer a fresh view of how nervous systems develop, adapt, and influence movement in bodies unlike those shaped by evolution.
