Rice University Engineers Use Light and AI to Reimagine Robotic Motion in Soft Arm Prototype
.webp?t=1755114198)
Once focused solely on electric products over the next several years the Roper Corporation is ramping up production of gas ranges previously made in Mexico. Photo courtesy of GE Appliances
HOUSTON, TX—Engineers at Rice University have developed a wire-free soft robotic arm capable of complex, real-time movements—powered and guided entirely by laser light and artificial intelligence. This breakthrough could pave the way for new approaches to controlling implantable medical devices and industrial robots designed to handle delicate or irregularly shaped objects.
In a proof-of-concept study published in Advanced Intelligent Systems, a team led by Dr. Hanyu Zhu, assistant professor of materials science and nanoengineering, demonstrated a light-activated robotic arm that bends, reaches, and navigates using a neural network-trained control system and an innovative polymer material. The system achieves motion without any onboard electronics, wires, or batteries.
“This is the first time a soft robotic system has been given real-time, reconfigurable, automated control using a light-responsive material,” said Elizabeth Blackert, lead author and recent doctoral alumna from Rice.
The robotic arm is made from a custom-engineered azobenzene liquid crystal elastomer—a smart material that contracts under blue laser light and relaxes in the dark. Unlike other photo-responsive materials, which often require UV light and long reset times, Rice’s elastomer responds within seconds using safer, visible light wavelengths.
“Our material bends toward laser light like a flower stem bends toward sunlight,” said Blackert. “That simple principle enabled us to create a highly adaptive, precise motion system.”
To control the arm’s movements, the team used a spatial light modulator to split a single laser into multiple “beamlets,” each targeting a different region of the robotic arm. Adjusting the intensity and position of these beamlets allows for real-time bending and flexing—similar to the fluid motion of octopus tentacles.
“This gives us virtually infinite degrees of freedom, which far exceeds the capabilities of traditional rigid robots,” said Dr. Rafael Verduzco, co-author and professor of chemical and biomolecular engineering.
Looking for quick answers on assembly and manufacturing topics? Try Ask ASM, our new smart AI search tool. Ask ASM
To achieve such precise control, the team trained a convolutional neural network (CNN) to recognize the relationship between light patterns and arm shapes. After analyzing a small dataset of light configurations and corresponding deformations, the AI model could accurately predict how to shape the laser to perform tasks like flexing, rotating, or reaching around objects.
The current prototype moves in two dimensions, but researchers are already planning future versions that can operate in full 3D using additional cameras and sensing technologies.
Soft robotics has long held promise in fields where traditional rigid robots fall short—especially in medicine and delicate manufacturing processes. The Rice team’s laser-guided, wire-free arm opens new possibilities for implantable biomedical devices, miniaturized surgical tools, and industrial automation systems that require non-invasive, reprogrammable motion control.
“By integrating smart materials, optical design, and AI, we’ve demonstrated a new framework for controlling soft robots,” said Zhu. “It’s a significant step toward building programmable, autonomous machines that are both safer and more adaptable.”
Looking for a reprint of this article?
From high-res PDFs to custom plaques, order your copy today!
