Hydrogels Mimic Neural Networks
“Ionic hydrogels can achieve the same kind of memory mechanics as more complex neural networks,” says first author and robotics engineer Vincent Strong of the University of Reading. “We showed that hydrogels are not only able to play Pong; they can actually get better at it over time.”
Inspired by a previous study where brain cells in a dish learned to play Pong through electrical stimulation, the researchers sought to explore whether simpler artificial systems could replicate similar feedback loops. “The basic principle in both neurons and hydrogels is that ion migration and distributions can work as a memory function that can correlate with sensory-motor loops in the Pong world,” explains corresponding author and biomedical engineer Yoshikatsu Hayashi.
Understanding Hydrogel Memory Mechanics
Hydrogels are complex polymers that become jelly-like when hydrated. The team used an “electro-active polymer,” a type of hydrogel that responds to electrical stimulation due to the presence of ions in the surrounding media. When electrically stimulated, these ions move, causing the hydrogel to temporarily change shape, a process likened to memory.
“The rate at which the hydrogel de-swells takes much longer than the time it takes for it to swell in the first place,” says Strong. “This means that the ions’ next motion is influenced by its previous motion, which is similar to memory.”
Teaching Hydrogels to Play Pong
To test whether this physical “memory” could enable the hydrogel to play Pong, the researchers connected it to a virtual game environment using electrodes. By sending the ball in a random direction and informing the hydrogel of the ball’s position through electrical stimulation, they observed the movement of ions to determine the position of its paddle.
As the game progressed, the hydrogel’s accuracy improved, resulting in longer rallies. Although the Pong-playing neurons in the earlier study reached their optimal ball-skills within around 10 minutes, the hydrogel required closer to 20 minutes to reach its peak performance.
Potential for New AI Algorithms
The success of this experiment suggests that hydrogels could represent a new kind of “intelligence” for developing simpler AI algorithms. The researchers plan to further explore the memory mechanisms within hydrogels and test their ability to perform other tasks.
“In our follow-up projects, we are thinking about how to extract the algorithm from the hydrogels that allows memory acquisition,” says co-author William Holderbaum. Strong adds, “We’ve shown that memory is emergent within the hydrogels, but the next step is to see whether we can also show specifically that learning is occurring.”
This research opens up exciting possibilities for the future of artificial intelligence and material science, where simple, non-neural systems might one day replicate complex cognitive processes.