Cynthia Sung is Making Robots for Everyone
To the average person, robots are still somewhat abstract: machines that operate in isolation in academic labs, factories or overseas for the military. The idea of the family robot, like the icon Rosie who bustled about doing everyday tasks on The Jetsons, is on par with the elusive jetpack, still something of a futuristic “maybe someday.”
Cynthia Sung, Gabel Family Term Assistant Professor in Mechanical Engineering and Applied Mechanics (MEAM), is looking to change that. Her research is making robot design and fabrication simpler, with the goal of getting to the point where anyone can make one.
“Robotics technology has huge potential to make people’s lives easier,” says Sung. “However, if robots are hard to use or hard to obtain, that potential will never be realized. We have already seen how robots have revolutionized manufacturing plants, space or deep sea exploration, and even disaster relief. It is not science fiction to say that they can similarly change our everyday lives.”
Robots integrate mechanical, electronic, and computational subsystems into physical devices that perform complex tasks ranging from autonomous navigation and manipulation to cooperative and social interactions. Sung became interested in robotics as a high school student when she followed the activities of the Mars rovers Spirit and Opportunity as they were launched and landed on Mars. “Around the same time,” says Sung, “my school started its first robotics team, and as a part of the team, I realized that robots were also technically interesting, and that they had broader applications than just space. That is probably when I first actively decided that I wanted to be a roboticist as a career.”
Through the development of a unique robotics platform, Sung aims to provide designers with intuitive computer-aided tools for creating customized robots and behaviors.
Drawing on research at the intersection of computational geometry, data-driven methods, and rapid fabrication techniques, Sung’s approach has a decidedly low-tech inspiration. Reminiscent of the sculptures produced through multiple folds and creases in paper, her robots look like they would be more at home on an arts-and-crafts table than a factory floor.
Sung earned a bachelor’s degree in Mechanical Engineering from Rice University and completed a doctoral degree in Electrical Engineering and Computer Science at the Massachusetts Institute of Technology, where her research focused on intuitive tools for robot design. It was at MIT Computer Science and Artificial Intelligence Laboratory where Sung’s “print-and-fold” paradigm of manufacturing took root.
“I have been folding origami since I was 10,” she says. “My mother taught me, and my favorite piece was always the flapping crane, because you could do more than just look at it once it was done, so I already knew intuitively that fold patterns had robotic potential.”
“When I entered the doctoral program, my advisor, Daniela Rus, and Cagdas Onal, who was a postdoc at the time, were just starting to look at how folding could be used for faster fabrication,” Sung recalls. “It seemed natural to me that I should combine my two interests, origami and robotics, to do foldable robot research.”
Sung’s approach involved computer tools, where users could specify the 3D geometry of a robot and how they would like it to move, and algorithms would plot out how such a robot could be built out of simpler components with various functionalities. For example, electronics modules were composed to create the entire control circuitry for the resulting robots, and fold patterns were composed to create valid fabrication plans.
At Penn, Sung is continuing her work in algorithms for synthesis and analysis of engineering designs from modular components. When considering the best place for the next step in her career, she was excited by Penn’s collaborative atmosphere. “Penn Engineering is a great place for this kind of interdisciplinary research,” she states. “I have found that the people here encourage making connections outside the traditional boundaries of a field, and I think my research plans will fit right in..”
In her research at Penn, Sung is advancing work on her robotics platform, seeing it as a way not only to expand accessibility to robotics, but also as a way to tackle manufacturing challenges. In a traditional setting, robotics assembly is difficult because many parts of different sizes and geometries need to be aligned and physically connected together. In contrast, a well-designed fold pattern automatically aligns during folding so that edges and faces of the 3D shape meet where they need to. “Folded robots present advantages in that they make use of purely planar fabrication techniques,” she says. “Up until the robot must be folded, all the components are flat. This significantly speeds up the fabrication process and also makes storage and transport much easier because you can pack a flat sheet instead of a complex 3D geometry.”
The group uses multiple different methods for creating their robots. The basic approach is to cut and score thin sheets of plastic or paper using a vinyl cutter, laser cutter, or even an old-fashioned pen and scissors. They also utilize a combination of 3D printing and folding to make more complex shapes, and more rigid shapes are fabricated using a layering approach that combines rigid faces of materials, such as plastic or metal, with flexible films. Electronics are currently added by hand, using off-the-shelf electronics modules for the robots’ simpler circuits and etching more complex ones directly onto their bodies. Upon folding, the electronic components align naturally due to the geometry of the pattern.
One of Sung’s short-term research goals includes developing a more intuitive user interface for designing a robot from start to finish. A current project, dubbed “Interactive Robogami,” is a design tool that aims to democratize the design and fabrication of robots, enabling people of all skill levels to specify and 3D print robots without the need for expert knowledge. The tool leverages a database of example robots that can be fabricated using the 3D-print-and-fold technique, and users compose parts from these examples to create new robot designs. Sung aims to enhance these kinds of interfaces so that they can account for additional variables like materials, the fabrication approach, environment and loading conditions, and robustness to fabrication and sensing errors.
Sung also aims to incorporate both mechanical and controls information into her interfaces. Existing CAD tools specialize in a particular subsystem, such as mechanical, electrical and software components, so design involves carefully switching between these tools so that any changes made in one subsystem are reflected in the others. She hopes to investigate how information about subsystems can be incorporated into a unified framework so that entire designs can be generated and verified in one tool.
In addition to user interfaces and design tools, Sung also wants to bring foldable robots to the point where they “can do real tasks.” Current work in foldable robot research mostly focuses on small-scale robots that require little power. “This is reflective of a lack of practical considerations, such as material properties, in folding theory,” says Sung. “My plan is to investigate fold pattern design methodology in the engineering context so that these robots can have real practical value.”
Above all, increasing access remains the driving force behind Sung’s research ambitions. Because advances in rapid fabrication methods like 3D printing are now more accessible, it follows that the hardware for robotics fabrication will eventually become more commonplace as well. With Sung working to make the design knowledge and methodology equally accessible, it is an encouraging time. “My hope is that eventually people who are not engineers and who do not have any engineering training can still specify, design, and take home their own custom robots,” she states. “I am very excited about the way that my field is moving.”