Equipping the Robots of Tomorrow with Expertise Honed at UW
GIX grad Tim Huff has a vision for the future of robotics
Transitioning from product management to engineering requires mastering an entirely new lexicon —almost a language of its own. At the University of Washington’s Global Innovation Exchange (GIX), Tim Huff (MS Tech Innovation 2024) not only learned a whole new lingo – he did so in Chinese.
Today, you will find Huff working as a robotics engineer at Groundlight AI in Seattle, but rewind to the early days of COVID-19, and you would have found him learning from instructors half a world away.
Harnessing GIX’s Dual-Degree with Tsinghua University
GIX, a collaboration between the University of Washington’s College of Engineering and the Foster School of Business, integrates design, robotics engineering, and business education. One of its standout programs is a dual-degree option for its students with Tsinghua University in Beijing, allowing students to earn two degrees: a Master of Science in Technology Innovation (MSTI) from the University of Washington and a Master of Science in Engineering (Data Science & IT) from Tsinghua in just 30 months.
Tsinghua University classes are typically conducted in English. However, as an undergraduate, Huff majored in Chinese Language and Literature. He studied at the Beijing Institute of Education, where his coursework was conducted entirely in Mandarin. This unique skill allowed him to take a machine-learning class taught in his Tsinghua professors’ native language.
Huff had planned to return to China for the Tsinghua portion of his education, but the pandemic forced a change of plans. Instead, his classes were conducted remotely. Huff studied in a language in which he is fluent but not native. It was challenging but rewarding.
“I speak Chinese, but there was a significant learning curve with the terminology around machine learning and other specialized aspects of my coursework,” he says. “Fortunately, my professors were excellent, and I was able to dedicate the extra time necessary to master the material.”
When Huff was ready to pursue the MSTI portion of his education, the pandemic had subsided. He studied onsite at GIX’s Bellevue campus, which includes a robotics lab where he got hands-on experience.
“I’m lucky,” he reflects. “The robotics coursework didn’t start until I was in Bellevue. It would have been very difficult to study robotics remotely.”
Enabling robots to work safely alongside people
At GIX, Huff honed the skills needed to secure a role as a robotics engineer at Groundlight AI, a Seattle technology company specializing in equipping machines with advanced visual capabilities.
Huff programs robots that can monitor a production line 24/7 and provide alerts on issues. They can “look” at a collection of parts to provide accurate counts and inventory. They can even answer human-language questions. Most importantly, they can do all this while operating safely in close quarters with humans.
Huff’s software enables assembly lines to operate more efficiently and helps operators detect anomalies in production. The word he uses most often to describe his work is not overly technical and more reassuring: “safety.”
“I’m most interested in how robots can safely operate alongside humans to make manufacturing easier,” he says.
Huff’s work at Groundlight builds on the projects he tackled as a student at GIX. He and his classmates used the program’s maker space to build a robot for one of the most hazardous jobs imaginable.
“For my launch (capstone) project, we put a robot underwater to inspect infrastructure,” he explains. “Typically, scuba divers are needed to check the condition of docks. But it’s dangerous. There are boats moving around and other hazards. So we put a robot down there instead, using computer vision to interpret underwater environments.”
Collaborations with GIX classmates
Huff collaborates closely with team members from various disciplines, including hardware experts and product managers. Thanks to his training at GIX, this multidisciplinary approach is familiar to Huff.
“GIX is very multidisciplinary,” he says. “There are designers, engineers, and people with different skill sets. Even within robotics, all twelve of us would work together. One team focused on navigation, figuring out how to best get from point A to point B. Another team worked on arm movement, determining how the robot picks things up. When you have a good team and everyone feels ownership over their part, those are the best projects.”
Tim Huff demonstrates Boston Dynamics’ “Spot” robotic dog. The device is remarkably agile, able to climb stairs, and has 360° perception.
Huff’s ability to work effectively with product managers is rooted in his prior career. Before becoming an engineer, Huff had a successful career as a product manager. He spent over five years in the field, including roles at Accruent and OrderPort, before deciding it was time for a change.
“I was working at a software company as a product manager, but I didn’t really know how to code,” he recalls. “I wasn’t a technical product manager. That’s why I went to GIX. I wanted to move in that direction, but I didn’t have the skills yet. I entered the program thinking I’d become a technical program manager. But after taking a few classes, I found I really enjoy coding and robotics.”
Bringing AI to next-generation robots
At Groundlight, his work utilizes both of those skill sets. He is at the forefront of bringing AI to robotics. Huff writes software for the next generation of robots, referred to as “cobots,” which work collaboratively with their human counterparts.
“A cobot is a robot that can operate in a factory alongside people,” he says. “For example, I wrote some code that allows cobots and other industrial robots such as Boston Dynamics Spot to ‘walk’ around the factory floor and take pictures of things and inspect them.”
Based on the cobot’s inspection, supervisors can pose binary questions to the device, such as “Is this manufactured correctly?” and have it answer yes or no.
Groundlight is working on expanding that dialogue between people and cobots. “We’re trying to make it easier for people,” says Huff. “If you want to use computer vision but you don’t have experts in that field, you can use Groundlight, and it’s as simple as asking that natural language question about the image. We are training the models, so every time you ask a question, you get the best possible answer.”
Bridging the gap between human and computer language isn’t the only way Huff is using his linguistic skills.
“I’ve actually been able to work with a few different Chinese robot companies here at Groundlight,” he says. “It’s especially great when I go to trade shows and Chinese people stop by—I can give them the demo in Mandarin!”
Learn more about the Global Innovation Exchange and its Master of Science in Technology Innovation (MSTI).
