UTSA’s ScooterLab received a $100,000 National Science Foundation grant to outfit store-bought electric scooters with computers and sensors to collect data that could help improve urban planning, traffic safety and even meteorological forecasting.
You could see students whizzing around UTSA’s main Northwest Side and downtown campuses on these scooters as early as next year — if researchers can land additional funding.
Murtuza Jadliwala, a UTSA assistant professor of computer science, conceived of the project in spring 2019 as he walked from the parking lot to his office and a student zoomed by on a scooter. He thought about scooter safety, and then he had an idea.
“These scooters are just riding on campus without being useful to anyone except the person who’s riding,” he said. “And I’m like, ‘Can we turn it around? Can we put these scooters to better use because we know that they’re going to be here?’”
His plan: create a network of campus scooters that students can use for free in exchange for the data the machines collect. The longer-term goal is for the network to expand to other higher-education institutions.
“We’re living in a data-centric world. Everything is data, especially artificial intelligence,” Jadliwala said. “So if you want to make advances in AI and machine learning, you need data.”
Jadliwala realized scooters have the potential to collect many types of “fantastic sensor data,” including motion, imagery and weather information.
“We are asking the government to help us build this network on one campus, and we will share the data with researchers around the world,” he said. “And anyone can effectively remotely deploy an experiment on our net.”
ScooterLab couldn’t turn to one of the commercial scooter companies — such as San Antonio-based Blue Duck, Lime, Byrd or Razor — because of the companies’ proprietary technology and their privacy and legal concerns. Beyond that, Jadliwala said commercial companies could skew the research.
In other words, they could use the collected data to learn how to boost their scooter rentals.
“The moment you include scooter companies, which have a bias motive, it taints your research,” he said. “So we have to be careful because we don’t want any conflicts of interest.”
The current NSF grant is to determine whether the technology is feasible.
Jadliwala enlisted Raveen Wijewickrama, a computer science doctoral student, as the project’s research assistant. The duo went to work building a computer setup that ties into the scooter’s electronic system.
“The first step is to build the technology and prove that, yes, that can be done — and that’s what we are doing right now,” he said.
Inside what looks like a toddler-sized backpack strapped to the front of the scooter’s handlebars, a Raspberry Pi computer motherboard is linked into the scooter’s electronics and a bevy of sensors that track global positioning, acceleration, humidity, pressure and temperature. It also includes a gyroscope.
The sensors provide researchers data about road conditions, patterns of travel, traffic, weather, ambient light, location, proximity, mobility, user privacy and more. Wijewickrama said the prototype scooter and computer costs about $1,000.
Student volunteers will be able to unlock the scooters, enabled with Wi-Fi and Bluetooth, via an app on their mobile devices. Future prototypes will have 4G and 5G wireless technology, according to Jadliwala, and the scooters will upload their data when they have connectivity.
ScooterLab currently includes an urban planning and micro-mobility specialist from UTSA’s College of Architecture as well as the university’s AI Consortium, a group of researchers.
“We are going to compare multiple different prototypes” of the scooters, Jadliwala said. “In the next six months, we are going to write a bigger proposal to NSF as a follow-up.”
Jadliwala said ScooterLab would need an NSF grant of roughly $1.5 million to deploy the first batch of scooters and build the back-end servers and hardware to support the network. He said that if ScooterLab gets the second grant, it could deploy the first 25 scooters in 2022 or 2023.
“Then from 25 we’ll scale up to 50, scale up to 100, and so on and so forth in the next two or three years,” he said. “One of the things we must all understand is micro-mobility technology, like any technology, is going to evolve, and we’ll need to evolve.”
©2021 San Antonio Express-News, Distributed by Tribune Content Agency, LLC.
