Electrical Engineering and Computer Science

Prof. Pei Zhang solemnly swears that he’s up to some good

In a project he calls the “Marauder’s Map,” Prof. Zhang uses machine learning-based data models, physics models, and heuristic models to turn physical structures into sensing devices.

Pei Zhang headshot Enlarge

Prof. Pei Zhang is a new faculty member in Electrical and Computer Engineering who’s looking to use the physical properties of devices and structures as sensors to discover information about surrounding physical elements, such as the location and movement of nearby people and animals. This project is dubbed the “Marauder’s Map,” a reference to the magical map in the Harry Potter series that reveals the whereabouts and movements of the inhabitants of Hogwarts.

Instead of magic, however, Zhang uses relies on machine learning-based data models, physics models, and heuristic models to improve the understanding of the physical systems. This work can benefit many fields including medicine, eldercare, farming, and smart buildings.

In this Q&A, we chatted with Zhang about his research, what he looks for in graduate students, his hobbies, and his special connection to U-M.

Tell us about your research.

We work on what we call, “Structures as Sensors.” We nicknamed it,“The Marauder’s Map,” because we’re literally building magic here.

Instead of using a small diaphragm in a microphone, we can use a humongous building or a fleet of cars as our physical element for the sensor. We install grophons to detect the vibrations of the physical building structure, but the sensors are so sensitive – their range is about 20 meters – we don’t need very many. The pattern of vibration we get can then tell us what’s happening. For example, when I was kid, I could tell whether it was my dad or my mom coming up the stairs based on their footsteps, so I knew whether I should be pretending to do my homework or not.

We're literally building magic here.

Prof. Pei Zhang

We can actually track and predict when an elderly person is about to fall, which is important for eldercare. We can track people’s heartbeats and breathing patterns, which are important metrics for medical professionals, especially when you have to monitor someone’s recovery after a major health condition, like an aneurysm or something.

We’ve also been working with the USDA folks to track and monitor pigs. So, mother pigs crush and kill a lot of their babies when they’re not in a good mood, but with our method, we can identify which ones are bad mothers, and we save a lot more of the litter that way. I remember having to crawl under a lot of defecating pigs to plant our sensors, so my students take great pleasure in calling it, “the sh***y project.”

How did the pandemic affect your ability to do this research?

We weren’t able to go to the Elder Care Center to collect data. One of our subjects is 100 years old, and we obviously want to keep her healthy, so that certainly put the data collection on hold. But, fortunately, we already had a lot of data before the pandemic started, so it actually made us focus more on the next stage. We were even more productive in some ways, because we were able to focus on writing our papers.

Why did you choose to come to U-M?

Michigan has a strong reputation in research, so it certainly is a top choice for any faculty or student. But I also have a personal connection. My grandmother was one of the few first Chinese students that came to the U.S. to study. She got her PhD here in physics or something, and then I think she stayed to do a bit of a research. So it feels a bit like coming home to me.

What got you interested in Electrical Engineering (EE)?

I took an EE class as an undergrad, and thought, hey, this is pretty interesting. I get to put things together and occasionally things might catch on fire and that’s actually kind of fun.

What do you enjoy most about your field?

The sensing world is really wonderful, because I get to work with so many amazing people and there’s such a wide variety of projects. I went to Kenya to help track zebras and got to hold a zebra’s head and put a collar on it while it was tranquilized. I work with kids with really severe genetic diseases that have life expectancies of 20 years, and we’re hoping to increase that to 30. And we want to improve care for the elderly.

It’s hard to imagine a more fulfilling career than this. Plus, I get to play at being Harry Potter.

It's hard to imagine a more fulfilling career than this.

Prof. Pei Zhang

What’s your favorite thing about teaching?

Witnessing a student go from someone who knows absolutely nothing to someone who knows as much or more than you is so fulfilling. You helped someone who will contribute to society and to the scientific world.

What qualities do you look for when selecting PhD students to work with on research?

It depends on the person. Some are going to be textbook smart, more theorist, and some are going to be better at the hands-on stuff. My job is to help them play to their strengths. They just have to be willing to learn. Curiosity and motivation to learn are the only two things I cannot teach.

Do you have any hobbies?

I scuba dive. I enjoy how low stress it is. You go into the water, float around, try not to move too much, and at the end someone pulls you back up on the boat.

One dive, I saw five whale sharks. I kicked really hard trying to get close and take pictures and I was so out of breath. But then I realized they weren’t moving fast, I was just really out of shape. I had to float there for like a whole minute for them to pass. It was like getting stuck waiting for a train. But they were so humongous, it was amazing.

More About Zhang

MA, PhD Electrical Engineering, Princeton University; BS Electrical Engineering, California Institute of Technology; Prior to joining U-M, Zhang was a faculty member at Carnegie Mellon University
Explore:
Division News; Electronics, Devices, Computers; Embedded Systems; Pei Zhang; Profile; Sensing and Sensors