Bretl Projects Merge Robotics and Neuroscience

Tim Bretl, from Beckman's Artificial Intelligence Group, merges disparate research areas in applications that aim to positively impact society.
Tim Bretl, from Beckman's Artificial Intelligence Group, merges disparate research areas in applications that aim to positively impact society.

Tim Bretl investigates two distinct areas of research: robotics and neuroscience. Yet Bretl’s research, enabled by a multidisciplinary team of students, merges the two disparate areas in applications that aim to positively impact society.

“It has been wonderful to work in these two areas because although they sound different, all my projects overlap. A lot of the same tools can be applied to all my projects,” said Bretl, who received his Ph.D. in 2006 from Stanford in aeronautics and astronautics and is now an associate professor of aerospace engineering. “I enjoy working with a diverse group of people who have expertise different from my own, and I’m lucky to have a great group of students who come to me with good ideas.”

One of Bretl’s four active projects is based in the area of robotic manipulation, specifically focused on how robots interact with flexible or deformable materials in manufacturing.

“An example of this work is automated manufacturing of cars,” said Bretl. “One thing that you have to do is install wiring in the cars. The wiring is generally bundled together and very flexible, and it’s very hard to automate its installation. We are doing a bunch of theoretical and algorithmic work to try to automate that process with a robot.”

“I’m trying to steer my group toward doing more applied work that has direct and beneficial impact on real people or society, which can be a challenge ... In particular, I try to do projects that are relevant to the community here.”-Tim Bretl

Bretl, a member of Beckman’s Artificial Intelligence Group, also works on a construction drone project, a collaboration between Mani Golparvar-Fard, assistant professor of civil and environmental engineering, and Derek Hoiem, assistant professor of computer science, and a member in Beckman’s Human Perception and Performance Group.

The group aims to use drones, mainly quadcopters, to automate inspection of large-scale construction sites. Currently, construction progress is monitored by workers who walk around the site, take pictures, analyze the data, and make decisions about resources based on how the construction is progressing.

With the drones, the process will happen autonomously, saving both time and money.

Bretl’s focus on the project is automating the process of data collection through the aerial robots. His team is developing methods of navigation and control that get the robots safely from one place to another on the site, take photos at ideal places, and guarantee that enough video is taken to support the visual analysis. Software then analyzes the video and makes recommendations for construction progress.

“There’s a huge need for this automation—it could potentially save millions of dollars in the construction industry,” said Bretl.

Bretl also works on the development of upper-limb prosthetic devices. He and graduate student Aadeel Akhtar are building a prosthetic hand that connects to the user with electrodes that read muscle activity (called an electromyographic [EMG]-based interface) and incorporates sensory feedback.

“Developing prosthetics is all about building effective communication with the device,” said Bretl. “There’s the EMG interface that translates muscle activity into commands for the device—so how the person tells the device what to do—and then there’s how the device tells the person what it’s doing—so that’s sensory feedback—and we do all of that.”
Researchers from a number of disciplines are involved in this project, including John Rogers, a member of Beckman’s 3D Micro- and Nanosystems Group and professor of materials science and engineering, who specializes in the development of the electrodes used for the EMG interface. The group also works with the Rehabilitation Institute of Chicago, where the prosthetics will ultimately be applied and tested on amputees.

Bretl and Akhtar are taking the prosthetics to the commercial marketplace soon—Akhtar won the Cozad New Venture Competition in April 2015 and is now funded by the University to advance PSYONIC, a start-up company that builds prosthetic hands for less than $1,000, compared to the $30,000 to $40,000 that current devices cost. The researchers hope that the application can improve the lives of people with amputations worldwide, especially in developing countries. 

The last of Bretl’s active projects, which are all funded by the National Science Foundation, involves brain-computer interfaces (BCI). His team is developing an interface that uses brain activity gathered from an electroencephalogram (EEG) to allow people to do anything from typing words to flying an aircraft.

Though the applications are endless, Bretl says one of the main objectives is to help people with severe motor disabilities interact more fully with the world around them.

Finding a way for BCIs to work would be a significant scientific accomplishment.

“You have to understand a lot of about brain function and how that’s reflected in EEG activity,” said Bretl. “So in thinking about how to get BCIs to work, it causes you to ask questions about the brain that tend to be a little different than what has traditionally been asked in neuroscience, and this could provide new insight for the scientific community.”

Robotics has always been a passion for Bretl, who, during his graduate studies at Stanford, worked on the algorithms that helped control the movement of the robotic legs that were used in Mars rovers. His interest in neuroscience began from collaborations with Todd Coleman, a former University of Illinois professor who specialized in network information theory.

“I mean, who doesn’t love robots? You figure out how to tell them what to do, and they do it—it’s a lot of fun,” Bretl said. “Even on the neuroscience side, a prosthetic hand is a robot that happens to be attached to a person. Robotics opens up a whole set of humanitarian applications and interesting science.”

Bretl plans to pursue more projects that have humanitarian applications by developing robotics that are commercially viable, including the automated drone monitoring of construction sites and prosthetic devices.

“I’m trying to steer my group toward doing more applied work that has direct and beneficial impact on real people or society, which can be a challenge,” said Bretl. “My own contributions in the past have tended to be fairly theoretical, but applied work is a lot of fun. In particular, I try to do projects that are relevant to the community here.”

His efforts to impact the community led him to a unique teaching position for the Education Justice Project, a program where University of Illinois professors teach college-level courses at the Danville Correctional Center to “demonstrate the positive impacts of higher education upon incarcerated people, their families, the communities from which they come, the host institution, and society as a whole,” according to the Education Justice Project web site.

Bretl taught a full course in robotics in spring 2013 and plans to teach another in spring 2016. He will cover topics taught in AE 353, a junior-level control theory course in the AE.

“It’s an incredible program and pretty unique in the country,” said Bretl. “It’s a pleasure to be part of it and a real opportunity for professors here.”

Bretl values the opportunity to positively impact society, whether it’s in his research or his teaching. He also received the College of Engineering Collins Award for Innovative Teaching and the AE Teacher of the Year Award, recognizing his excellence as a professor in undergraduate courses.

“Faculty are given a lot of freedom to do what we consider to have value. Projects that matter to people here in my community—both inside and outside the university—are what I want to spend my time on.”