Sensors, sensors everywhere

Patwari, who holds a dual appointment as a professor of electrical & systems engineering and of computer science & engineering, is an expert in using radio waves, the signals used by wireless devices, to develop sensors that can be used in applications including medical devices or to detect motion in a home. He and his lab, the Sensing and Processing Across Networks (SPAN) lab, use the same range of frequencies used in cell phones and other wireless devices to create these sensors.

 

Nearly eight years ago, Patwari, who joined the McKelvey School of Engineering at Washington University in St. Louis in January 2019 after 12 years at the University of Utah, and his team were working on a project using radio waves to determine which rooms were occupied in a home so that lights could turn on and off automatically in response to motion. A student alone in the lab saw changes in the measurements of radio waves although there was no movement in the lab. On a hunch, he held his breath, and the measurements stayed constant — the device was detecting the radio waves from his breathing.

“We just happened on to it,” said Patwari, a native of Cleveland, Ohio. “It was rare to observe that on lab equipment, so we had to take some time to figure out how to make a commercial device that would reliably record breathing no matter what your position was or what else was going on.”

Based on this, Patwari’s lab has developed sensors to measure breathing and pulse rate that do not need to be connected to the body. Their sensors can be placed on opposite sides of a bed to determine changes in the radio waves due to breathing changes and determine a pattern over time. Now, he and his lab are working on wireless protocols for a router that would secure that breathing and pulse rate data.

“The encryption and security technologies that have emerged for Wi-Fi and cell phones don’t help at all because they encrypt the data being sent,” he said. “Vital sign sensing doesn’t require the data because it observes changes occurring to the radio waves. We think that by having an access point insert random changes into its transmitted signal we can allow people near that access point to keep information about their vibrations and health private.”

Patwari also is working to get his sensors into the market through his association with startup Vita Sensors, which has made a prototype medical breathing monitor. He also was director of research for Xandem Technology from 2010-2019. Before earning a doctorate from the University of Michigan, he also spent two years as a research engineer at Motorola Labs, which obtained several patents for his work. In all, he has a hand in 14 patents, some of which have been licensed by Vita Sensors, Xandem Technology and Texas Instruments.

Patwari and students in his lab also are developing sensors for football and other sporting helmets that might warn the wearer of an impending collision. He and his team are working with mechanical engineers and biologists to test it out.

“The idea is if someone is going to get hit from behind, they should get a warning,” he said. “Maybe it’s not long enough to change the game, but enough that they would tense up their muscles and reduce the angular acceleration to the head. If we can achieve that goal, it could reduce traumatic brain injury.”

Over the course of his career, Patwari has been awarded, together with his collaborators, more than $24 million in research grants from such agencies as the National Science Foundation (NSF), including a CAREER Award given to promising early-career investigators; the National Institutes of Health; the Army Research Office; and the University of Utah Research Foundation. The largest grant he and his team are working on is one for the NSF- and PAWR-Industry-Consortium-funded Platform for Open Wireless Data-driven Experimental Research (POWDER), a testbed for experimenting on the future of wireless networking in a living laboratory composing 10 square kilometers in Salt Lake City. There are fixed stations as well as buses and utility vehicles that carry programmable wireless devices around the city.

This testbed allows users to log on to a base station and user device from home, write code to program them and have them operate as their code describes,” Patwari said. “The idea is that when people are making the next-generation developments for 5G or 6G, they can test their idea in the real world rather than in a lab.”

At WashU, Patwari designed a graduate-level course for the Fall 2020 semester titled “Equity and Fairness in Estimation and Classification,” about fairness in machine learning. The course has given Patwari the opportunity to have meaningful discussions with students about issues of race, gender, class and disability, and how algorithms have undesirable and unfair effects on society.

“We are looking at how machine learning algorithms learn from society’s inequities,” he said. “They might be trained using a set of data that includes only the dominant group or has labels generated by a biased world, and these biased algorithms then contribute to further inequities in policing, criminal justice and real estate. All of these inequities are visible in St. Louis, so it is the perfect place to teach this material.”

Patwari also is chairing a faculty recruiting committee for the university’s race and ethnicity cluster hire initiative, which seeks to hire 12 new faculty members engaged in research on the manifestations of race in society. Patwari’s committee is seeking a faculty member engaged in data science and racial equity who will have an appointment in the Division of Computational & Data Sciences and in McKelvey Engineering’s Computer Science & Engineering department.

Patwari’s new course has been developed with consultation from his spouse, Cassie Power, associate director for faculty & academic engagement with the Gephardt Institute for Civic & Community Engagement, who earned a doctorate in women’s studies and social psychology and has 12 years of experience in teaching with a critical framework. Her contribution has ensured that the course provides the necessary critical race and feminist theory to be able to see algorithms in the larger societal context in which they operate.

“I’m really fascinated by all of these topics and to learn the kinds of tools I can use in engineering to make some contribution,” he said.

Back to Engineering Momentum