Artificial Intelligence (AI) and robotics are more prominent in modern society than ever before. This is thanks to their ability to improve workflow and communication, and to advance our technological abilities. However, scientists now face the challenge of finding rapid ways to adapt robots to external stimuli, so they respond properly and move more fluidly through their environment. To achieve these goals, scientists study complex systems of brain cells that send signals to each other to communicate, called neural networks.
A team of researchers from Cornell University sought to improve the limitations of robotics that computer-based programs struggled to overcome in the past. These include short life spans, intensive procedures, and low environmental reactivity. They tested whether a system called biohybridization could improve neural networks. This technology uses living materials like tissues or cells in conjunction with synthetic materials. They hypothesized that biohybrid robots would react more quickly to unpredictable situations and solve problems more naturally than computer-based programs.
Researchers in the past have used neural networks based on animal or plant cells to improve robots’ movement and reactivity to the environment. However, difficulties can arise due to the intensive care needed to keep these cells healthy in an artificial setting. In particular, animal cells need a healthy metabolism, which requires diligent care and careful monitoring to avoid contamination and cell damage.
These researchers tested neural networks based on a more resilient non-animal system, fungi. Fungi use electrical signals to transmit information, similar to the neural signals animals use to communicate. In doing so, fungi create extensive underground threads that connect into a web-like structure, called mycelial networks.
Fungi use mycelial networks to transport nutrients, sense electrochemical signals, and communicate environmental stimuli to other fungi. Fungi are also hard to contaminate in cell cultures due to their high survival rate and resilience. Unlike animal cells which need constant attention, fungi are more cost-effective due to their less extensive culturing process and rapid growth.
The researchers built 2 robots, one shaped like a starfish that could move its arms independently, and another that only moved forward and backward. They selected the king oyster mushroom fungus to study because it’s non-toxic and grows fast in the laboratory. They set up the fungi so their mycelial networks grew directly on specific nodes of the control board of the robot’s interface. Then they measured natural electrical signals produced by the networks and watched them react to stimuli.
To test how the fungi affected the robots, the scientists allowed the fungal network to grow toward the robots’ electrical nodes and measured the bioelectrical signals it produced. They compared their results with bioelectrical signals from another sample of fungi that had not grown inside a robot. They found that the fungi grown inside the robots produced more electrical impulses than the free-growing fungi. According to the researchers, these observations indicated that the network was active enough to control the robot’s interface and function effectively.
Next, the team tested which external stimuli triggered the fungi. They exposed the fungi to different types of light and found that it reacted negatively to the Sun’s UV light and to blue light similar to that emitted from screen devices.
Once they had created the robot interfaces, grown the mycelial network, and found the best source of stimuli, the researchers set out to test how the robots reacted in different scenarios. They examined the robots’ response to UV light to test their basic movement. They placed each robot in a box with a UV light outside. When the robots left the box and sensed the light, they responded by returning to the box to avoid the light. This response demonstrated the fungi’s ability to control a biohybrid robot using only their natural electrical signals.
The scientists concluded that fungi biohybridization could advance the field of robotics due to the resilient nature of fungi compared to animal or plant cells. They suggested that researchers could use fungi in more advanced robots since mycelial networks can sense chemical shifts and communicate through complex pathways. They proposed that this milestone creates new routes for robots to adapt to their environment in the field and provide their users with more reliable sensory information.
