Tiny, soft robot flexes its potential as a life saver

UNIVERSITY PARK, Pa. — A tiny, soft, flexible robot that can crawl through earthquake rubble to find trapped victims or travel inside the human body to deliver medicine may seem like science fiction, but an international team led by researchers at Penn State are pioneering such adaptable robots by integrating flexible electronics with magnetically controlled motion. 

Soft robotics, unlike traditional rigid robots, are made from flexible materials that mimic the movement of living organisms. This flexibility makes them ideal for navigating tight spaces, such as debris in a disaster zone or the intricate pathways of the human body. However, integrating sensors and electronics into these flexible systems has posed a significant challenge, according Huanyu “Larry” Cheng, James L. Henderson, Jr. Memorial Associate Professor of Engineering Science and Mechanics at Penn State. 

"The biggest challenge really was to make it smart," said Cheng, co-corresponding author of the team’s study published in Nano-Micro Letters. "For most applications, soft robotics have been a one-way communication system, meaning they rely on external control to navigate through complex environments. Our goal was to integrate smart sensors so these robots could interact with their surroundings and operate with minimal human intervention." 

A principal factor in making these robots smarter lies in the integration of flexible electronics, which enables their key features. 

"We wanted to design a system where soft robotics and flexible electronics work together seamlessly," Cheng said. "Traditional electronics are rigid, which makes integration difficult. Our solution was to distribute the electronic components in a way that preserves the robot’s flexibility while maintaining robust performance." 

Cheng and his team shot videos of the robots in action, capturing their dynamic behavior as they crawl and roll into a ball to move along a simple course. The robots move using hard magnetic materials embedded in their flexible structure, which allows the robots to respond predictably to an external magnetic field. By adjusting the field’s strength and direction, researchers can control the robots’ movements, such as bending, twisting or crawling, without onboard power or physical connections such as wires.  

A major hurdle in developing this technology was figuring out how to keep the flexible electronics from hindering the robot’s movement.  

"Even though we designed the electronics to be flexible, their stiffness is still hundreds to thousands of times greater than the soft robotic material," Cheng said. "To overcome this, we distributed the electronics across the structure, reducing their impact on movement." 

Another challenge was blocking unwanted electrical interference, which can disrupt how an electronic device or system works. This interference comes from outside sources, like other electronics or wireless signals. Such interference would hinder movement and affect sensor performance.  

"Magnetic fields are crucial for controlling motion, but they can also disrupt electronic signals," Cheng noted. "We had to carefully design the electronic layout to minimize these interactions, ensuring that the sensors remained functional even in the presence of strong magnetic fields." 

With the magnetic interference minimized, the robots can be guided remotely using electromagnetic fields or handheld magnets — which limits the human intervention they need. Additionally, integrated sensors allow them to react autonomously to environmental cues. In search-and-rescue, for example, they are smart enough to navigate debris by detecting heat or obstacles. In medical applications, they might respond to pH changes or pressure, ensuring precise drug delivery or accurate sample collection.  

The next step for Cheng’s team is to refine the technology for such applications — including creating a “robot pill.” 

"One of the most fascinating potential applications is in implantable medical devices," said co-author Suk-Won Hwang, associate professor at the Graduate School of Converging Science and Technology, Korea University. "We’re working on miniaturizing the system to make it suitable for biomedical use. Imagine a small robotic system that could be swallowed like a pill, navigate through the gastrointestinal tract and detect diseases or deliver drugs precisely where they’re needed." 

Such technology could provide a less invasive alternative to traditional diagnostic procedures, like biopsies, gathering data directly from the patient in real time, according to the researchers. 

"With integrated sensors, these robots could measure pH levels, detect abnormalities and even deliver medication to precise locations inside the body," Cheng explained. "That means fewer invasive surgeries and more targeted treatments, improving patient outcomes." 

Cheng said he also envisions future applications in vascular treatments.  

"If we can make these robots even smaller, they could be injected into blood vessels to treat cardiovascular diseases or deliver medication directly to affected areas," Cheng said. "That would open up entirely new possibilities for non-invasive medical treatments." 

While the team hasn’t yet given these robots an official name, Cheng said they are open to suggestions.  

"That’s a good suggestion," he said with a laugh. "Maybe we should get the public involved in naming them." 

Readers are invited to submit naming ideas to Cheng at huanyu.cheng@psu.edu.  

Along with Cheng and Hwang, other authors of the study from Penn State include Bowen Li, research assistant in engineering science and mechanics, and Ankan Dutta, doctoral student in mechanical engineering who is also affiliated with the Center for Neural Engineering. Joong Hoon Lee, Gwan-Jin Ko, Tae-Min Jang, Won Bae Han, Sueng Min Yang, Dong-Je Kim, Heeseok Kang, Jun Hyeon Lim, Chan-Hwi Eom and So Jeong Choi, KU-KIST Graduate School of Converging Science and Technology, Korea University; and Sungkeun Han, Semiconductor R&D Center, Samsung Electronics Co., also contributed to the paper.  

The National Research Foundation of South Korea and the Korea Institute of Science and Technology supported this research.