The CITRIS principal investigator and assistant professor of mechanical engineering takes inspiration from nature and uses machine learning to create more efficient materials.
“CITRIS is really unique because it facilitates new collaborations in the UC system,” said Grace Gu, an assistant professor of mechanical engineering at the University of California, Berkeley, and a principal investigator (PI) at the Center for Information Technology Research in the Interest of Society and the Banatao Institute (CITRIS).
“I wanted to explore a new area of research, and a CITRIS Seed Award opened doors to new partnerships and opportunities.”
Gu, who joined the UC Berkeley faculty in 2018, leads a research group that uses artificial intelligence (AI), advanced computational analysis and inspiration from biological structures to engineer better materials and improve additive manufacturing techniques.
Gu has been interested in finding meaningful applications of her work — a core tenet of the CITRIS mission — since her days as an undergraduate mechanical engineering major at the University of Michigan, where her senior capstone team designed and patented a minimally invasive skin biopsy device to help make an uncomfortable process easier for patients and clinicians.
Gu’s research in bioinspired materials began in graduate school at the Massachusetts Institute of Technology (MIT). During her doctoral studies, she investigated the structure of bone, alligator skin and different types of seashells, including the exceptional architecture of conch shells.
As Gu explains, seashells may look unassuming, but conch shells are composed of multiple intricate layers that make it more difficult for predators to shatter them to expose the fragile organism inside. Despite being made up of jelly-like proteins and chalky minerals, these mollusk-manufactured materials are extraordinarily tough and resistant to impact.
“Natural materials are really interesting because through billions of years of evolution, they become optimized for their environment and their needs,” she said.
Gu’s relationship with CITRIS and the Banatao Institute began just a few years after her arrival in Berkeley. She heard colleagues mention the CITRIS Seed Funding Program, which supports early-stage, multicampus projects that have the potential to address major societal challenges.
Gu saw an opportunity to dive deeper into research inspired by another structurally superlative sea creature: the fast-swimming mako shark. She reached out to fluid dynamics expert Stephen Robinson, a professor of mechanical and aerospace engineering at UC Davis, the director of the NASA-funded Habitats Optimized for Missions of Exploration (HOME) Space Technology Research Institute and a former astronaut, to develop a proposal.
Their project received a 2021 CITRIS Seed Award. The team aims to reduce air drag on, and thus the fuel needs of, commercial airplanes by replicating the aerodynamic nature of mako shark skin. Though it may look smooth, shark skin is covered in microscopic scales shaped like teeth, called dermal denticles. These denticles cut through the water, helping the sharks swim faster.
However, as Gu says, “Nature has answers to nature’s problems, which aren’t necessarily the same problems that we face in our modern world.” She is using AI to identify what configuration of the denticles will best suit human needs, within human manufacturing limitations.
She compares this bioinspired, computationally optimized design process to figuring out how to build a house with a pile of toy building blocks of different shapes and sizes, without any instructions. Brute-forcing a solution could take years, but AI can theoretically fast-track evolution and find a solution in a much shorter period of time.
Gu and her research team “translate” material structures into the binary code that computers can understand and then put them through an optimization algorithm to determine the best structure for given usage requirements, such as lower air drag or higher heat resistance, as well as manufacturing constraints, such as 3D-printer resolution.
This machine learning technology is similar to how AI can be trained to decipher whether an image shows a cat or a dog by detecting patterns in color and texture, except that it is now designing three-dimensional models by recognizing structural features.
The team has been working to reshape shark denticles to a configuration that is well suited to cover the panels of commercial airplanes. Then, under Robinson’s guidance, they test the materials and validate their prototypes in the UC Davis aeronautical wind tunnel.
“The CITRIS Seed Funding program really helped to catalyze this collaboration,” Gu said.
While the Seed Award focuses on applications in aviation, she notes that denticles could also be designed to coat automotive vehicles, wind turbine blades, gas pipelines, even swimsuits — anything that would benefit from reduced drag.
Beyond her CITRIS-supported efforts, Gu’s group is pursuing several other research avenues with biological underpinnings. One area of investigation, supported by the UC Berkeley Bakar Fellows Program, integrates Gu’s interests in additive manufacturing and biomedical applications. As a Bakar fellow, she is working to help people heal from ligament ruptures by “bioprinting” custom replacement parts from collagen. She aims to produce a prototype within three years.
The wings of owls have also offered inspiration. Unique among birds for their silent flight, which helps them approach prey without detection, owls have serrated feathers that break down air turbulence to reduce noise. Gu is hoping to apply lessons learned from these feathers to the design of drone propellers for use in urban settings.
“If you’re thinking about delivering packages, you don’t want the noise to disrupt the neighborhood, right?” she said.
In a collaboration with MIT Lincoln Laboratory, Gu and team are planning to combine features from both shark denticles and owl wings in uncrewed underwater vehicles, or UUVs, to lessen drag, reduce noise and thus optimize their ability to operate in stealth.
She is also working on a multi-university project funded by the National Science Foundation’s Designing Materials to Revolutionize and Engineer our Future (DMREF) program to combine structural features from octopuses, chemical properties from mussels and AI to engineer adhesive devices that can be easily removed and reapplied. This switchable adhesive technology could help make prosthetics fit better, robots grip better and pick-and-place manufacturing processes run more efficiently.
Although numerous and wide-ranging, Gu’s efforts in the laboratory maintain focus on applications that improve people’s lives.
“For research, it’s important to think about why you’re doing it, and what steps you have to take in order to achieve the goals you’ve set,” said Daniel Lim, a doctoral student in Gu’s group. “Professor Gu mentored me well on how to define my research process. I’m very lucky to have her as my advisor.”
Gu is also dedicated to making her research more accessible in the hopes of inspiring the next generation of scientists and engineers.
She has created outreach activities for elementary and middle schoolers with UC Berkeley Girls in Engineering (GiE) and the Society for Women Engineers. In January 2023, Gu and her group organized a hands-on workshop for a dozen high school and undergraduate students. The students built a 3D printer from scratch, sharpened their computer-aided design (CAD) skills and tackled a challenging design project that made use of the new machine.
Gu is also a mentor for the UC Berkeley Transfer-to-Excellence (TTE) Summer Research Program, a nine-week research internship that brings California community college students into the labs of UC Berkeley science and engineering professors.
Here she made use of another CITRIS resource: A recent TTE intern completed his research project with the help of the CITRIS Invention Lab, which allowed him to quickly manufacture 3D-printed prototypes. With the assistance of Daniel Lim, a former Invention Lab superuser, 2022 intern and Fullerton College student Jacob Lopez was able to create physical models of semiconductors — something almost unheard of given the short timespan of the program.
“It’s hard to achieve physical outcomes in that period of time, but thanks to Chris Myers and Dan Chapman, Jacob was able to actually use the equipment to fabricate a complete experimental setup,” Lim said. “It really boosted up the thought process behind prototyping in research.”
In 2022, she gave a well-received TEDxBerkeley talk, which has helped to explain the value of bioinspired materials, and the potential of AI-optimized design, to tens of thousands of viewers.
She will also moderate a panel on sustainability in advanced manufacturing at the 2023 EDGE in Tech Symposium in March, in which a group of industry experts will discuss how information technology can help balance the competing drivers of economic and environmental stability in product development and distribution.
“The CITRIS program has unlocked a world of possibilities,” Gu said. “I am excited to collaborate with students and partners to pave the way towards a sustainable future through innovative solutions.”