The comforting embrace of a knit sweater isn’t just for warding off the winter chill anymore. Researchers from Carnegie Mellon University’s Robotics Institute have repurposed this cozy concept to open new doors in human-robot interaction.
The team developed a technology called RobotSweater – a machine-knitted textile “skin” that enables robots to perceive touch and pressure (see image here).
“Through this technology, we aim to make the robot smarter during its interaction with humans,” said Changliu Liu, an assistant professor of robotics at the School of Computer Science.
The process of creating the RobotSweater draws parallels to the flexible, adaptive nature of knitting. Much like how a skilled knitter can shape yarn into any form – be it a sock, hat, or sweater – the RobotSweater’s textile skin can be tailored to snugly fit robots with uneven three-dimensional surfaces.
Professor James McCann, whose recent research was focused on textile fabrication, said: “Knitting machines can pattern yarn into shapes that are non-flat, even curved or lumpy. This capability sparked the idea that we might create sensors to fit over such unconventional robotic forms.”
Once fabricated, the RobotSweater lends a new dimension of sensitivity to robots. It equips them to “feel” human touch – a feature of utmost importance in industrial settings where safety is key. Existing solutions for human-robot interaction detection often resemble rigid shields, which can’t cover the entire robot body due to the necessity for certain parts to deform.
“The RobotSweater, on the other hand, allows for full-body coverage, thereby enhancing the robot’s ability to detect any possible collisions,” explained Liu, who specializes in the industrial applications of robotics.
The RobotSweater’s textile skin consists of two layers of conductive yarn made with metallic fibers for electrical conductivity. In between these layers, there is a lace-patterned, net-like layer. When someone applies pressure to this fabric, it closes an electrical circuit, enabling the sensors to read the contact.
Professor Wenzhen Yuan, director of the RoboTouch lab, clarified how this works: “The force pushes together the rows and columns to close the connection. If there’s a force through the conductive stripes, the layers would contact each other through the holes.”
While figuring out the design and knitting patterns posed a challenge, the team also grappled with the intricate task of connecting the wiring and electronics to the soft textile.
“It entailed a lot of meticulous physical prototyping and adjustment,” said McCann. The team succeeded by fastening wires around snaps affixed to the ends of each stripe in the fabric. These snaps are not only cost-effective but are also an efficient solution for hobbyists interested in creating textiles embedded with electronic elements, known as e-textiles.
In order to withstand stretching without damaging the yarn, the team needed a robust yet flexible connection method. McCann added that they also contemplated the use of flexible circuit boards.
Once RobotSweater is fitted onto a robot, it can sense the distribution, shape, and force of contact. This sensitivity exceeds the precision and effectiveness of conventional visual sensors that robots typically rely on.
“The robot will move in the way that the human pushes it, or can respond to human social gestures,” said Yuan. The team’s research showcases how pushing a RobotSweater-adorned companion robot can guide its movement or direct it to turn its head. Similarly, on a robot arm, RobotSweater can interpret a hand’s push to guide its motion, while a firm grasp signals it to open or close its gripper.
In future research, the team plans to explore the potential of programming reactions from touch-screen-like swipes or pinches.
The researchers – including PhD students Zilin Si and Tianhong Catherine Yu from SCS, along with visiting undergraduate student Katrene Morozov from the University of California, Santa Barbara – are ready to present the RobotSweater research paper at the 2023 IEEE International Conference on Robotics and Automation (ICRA).
The collaboration began simply enough – with three faculty members discussing the idea over lunch. However, the diversity in their specializations facilitated the realization of the RobotSweater project.
“There was a person thinking about fabrication, another about robotics integration, someone focused on sensing, and another working on planning and control,” explained McCann. “It’s truly remarkable to have this full stack of people addressing each concern.”
Thus, by transposing the very qualities that make a knitted sweater comfortable and easy to wear, the Carnegie Mellon research team has created a textile “skin” for robots.
This innovation, the RobotSweater, not only enhances the safety of human-robot interaction but also promises a more intuitive and human-like robotic response. The potential implications of this development, from industrial applications to personal robotics, are both exciting and vast.
Robots have come a long way since their conception, and they continue to evolve at an astonishing pace. These technological marvels are increasingly interweaving with our lives, reshaping various aspects of human society, and influencing the Earth at large.
Industrial robots have seen significant advancements over the past decades. They’ve become more agile, precise, and capable, transforming manufacturing processes worldwide.
Automated machines now perform complex tasks, from assembling intricate electronics to welding and packaging, with minimal human intervention. They’ve improved efficiency and productivity, while reducing human exposure to potentially dangerous tasks. The advent of collaborative robots, or cobots, has allowed safe and efficient human-robot collaboration in the workspace.
Robotics technology has made considerable strides in healthcare. Surgical robots, like the Da Vinci Surgical System, have enabled minimally invasive procedures with increased precision and reduced recovery time. Robotic prosthetics and exoskeletons are offering new possibilities for people with disabilities, aiding mobility and dexterity. Autonomous robots are even being used for patient care, medication delivery, and disinfection in hospitals, especially highlighted during the COVID-19 pandemic.
Robots are increasingly being used in environmental monitoring and conservation efforts. Unmanned aerial vehicles (UAVs or drones) and underwater robots help track wildlife, monitor pollution, and map landscapes in inaccessible areas. Robots are also being developed to recycle waste, clean up oil spills, and perform tasks that help reduce our environmental footprint.
The integration of AI in robotics has led to a new breed of intelligent machines. AI-powered robots can learn, adapt, and make decisions, making them more autonomous. They are capable of natural language processing, visual perception, and even understanding human emotions. This advancement has significant implications in fields like elder care, where robots can provide assistance and companionship.
Social robots, designed to interact with humans in a personal and emotive manner, are becoming more prevalent. From therapeutic robots like Paro, designed to comfort elderly dementia patients, to educational robots helping children with autism, these machines are forging new paths in human-robot relationships.
Robots have been indispensable in exploring the cosmos. Mars rovers like Perseverance and Curiosity are conducting groundbreaking research on the red planet, and robots are planned for future missions to the moon, asteroids, and beyond.
The impacts of these robotic advancements on humanity and Earth are profound. They’re revolutionizing industries, healthcare, environmental management, and space exploration, among other fields. However, these advancements also come with challenges, such as job displacement due to automation and ethical concerns around AI and social robotics.
As we continue to navigate this exciting frontier, it’s crucial to consider these implications and strive for a future where technology serves to augment human abilities and better the world, while minimizing potential downsides.