Research professor Marjorie Zielke PhD’07 and a company founded by computer science senior Ashlesha Nesarikar have both won the city of Richardson’s Smart Gigabit Communities development challenge.
The challenge is part of US Ignite’s encouragement for participating cities to create Smart City applications addressing community issues such as public safety, transportation and health care.
Zielke and Plano Intelligence Inc. (PI) each received $5,000 for winning the proposal portion of the challenge. Both will be eligible for a second grant of the same amount when they demonstrate a developed prototype of their application within six months.
Using AI for iNotify
Nesarikar and her team took first place for PI’s proposal for a smartphone-based application called iNotify. The app uses a proprietary artificial intelligence system to locate individuals with a weapon at sensitive locations and notify police, administration and other interested parties before an active-shooter situation may develop.
From left: Members of Plano Intelligence Inc. are Lirit Fuksman, Nikhil Gupta, Ashlesha Nesarikar, Siddharth Naik, Mustafa Sadriwala and Alan Liao.
The idea for her team’s proposal started inauspiciously. The family dog kept stealing food from the countertop when nobody was looking. Unable to ever catch the dog in the act, Nesarikar turned to her technological strengths and developed a series of sensors that, using artificial intelligence, detect the dog’s approach and alert her via a remote system.
Later, a more sobering incident drove home the need for the app.
“The first time I thought about it seriously was after the stabbing at UT Austin last summer,” Nesarikar said. “My friend was texting me, and I got the whole experience secondhand.”
She began developing a proposal for her AI-driven system built on her earlier framework that was eventually submitted for the Smart Gigabit Communities contest.
The concept for iNotify is that it analyzes security footage in real time via both deep neural networks and recurrent neural networks. Information on dangerous weapons, such as guns and knives, are fed to an AI system that can then identify suspects, and alert police departments and emergency responders before an attack ever starts.
“It really became evident to me about how important it was after the events at Parkland High School,” Nesarikar said. “It actually happened while we were presenting our app proposal for the contest. We didn’t know at the time, but it turns out the judges did.”
The other members of her team — biochemistry senior Lirit Fuksman, computer science freshman Siddharth Naik, computer science sophomore Alan Liao and computer science junior Mustafa Sadriwala — will use the grant to create the front-end user interface, along with more sophisticated iterations of the core AI.
Nesarikar said feedback will be crucial to deciding what features need to be added, such as customized alerts for parents with children at multiple schools across a city. Nesarikar and her team are excited to continue what they see as necessary work.
“The app won’t just improve lives; it will save lives,” Fuksman said.
Emotive Virtual Reality Patient System
The second winning proposal was Zielke’s plan to add new communications capability to the Emotive Virtual Reality Patient System App that she and her team developed at the Center for Modeling and Simulation/Virtual Humans and Synthetic Societies Lab.
The team is using a Microsoft HoloLens so medical students can develop interpersonal skills with patients by interfacing in real time with virtual personas developed by Zielke’s team.
When the headset is activated, the wearer can see the virtual patient sitting or standing in an examining room. They can then prompt the patient with a number of questions devised to help the doctor investigate the patient’s symptoms. Furthermore, the interactions allow medical school students to develop social skills necessary for working with patients in a daily setting.
“With these resources, we have been able to create one of the first networked augmented reality systems in the country as well as explore how to use the US Ignite network to develop the ‘brain’ of our virtual patient,” Zielke said.
The foundation funding for this project is a collaboration with The University of Texas Southwestern Medical Center and was provided by the Southwestern Medical Foundation.
The grant money will allow Zielke’s team to expand the network capabilities of the existing personas and explore new applications for high-speed networks. Zielke’s team has presented its prototype at several conferences in the U.S. and Canada, and attended National Science Foundation workshops.
“The capabilities of the US Ignite network, along with very unique UT Dallas capabilities such as expertise in 3D character development will be critical to our goal of creating world-class human emulation,” she said.
