If “two heads are better than one,” then sharing mobility data with stakeholders and even the public at large has the potential to catalyze important innovations in mobility technology. That’s the premise behind a number of projects in metro Detroit that are collecting and sharing information about traffic infrastructure and the movements of cars, bikes, scooters, and other forms of transportation to improve efficiency, safety, and equity.
Steve Remias describes southeast Michigan as a “global leader” in collaborating on mobility solutions, but says data sharing in particular is just beginning to catch on. Remias is an assistant professor at Wayne State University’s (WSU) College of Engineering and a member of WSU’s CAR Lab, which focuses on developing technology to enable connected and autonomous driving.
Remias says CAR Lab and others at WSU are already collecting and sharing mobility data through a number of projects. He notes that companies everywhere in all fields, not just mobility, are collecting “terabytes on terabytes of data,” but aren’t using it to the fullest.
“For a really long time, we’ve been data-rich, but information-poor,” he says. “That’s where having shared data sets can tease out more of those information messages, as opposed to having billions and trillions of rows of data. We have all of this data, but to make it more valuable, we need to convert it into something agencies and academics can (use to create) some action items and provide value to the public.”
Remias praises the Southeast Michigan Council of Governments (SEMCOG) for rethinking the data silo approach.
“SEMCOG has a very nice traffic data-sharing platform, using GIS, crash data, census data and other kinds of data on their website,” he says. “It’s a really nice platform for city planners and academics to use.”
“There are still data silos in this mobility space where people collect their data and keep it close to the vest,” he says. “However, in my 10 years of working in this area, individuals, academics, agencies, and industry have all started to realize that data sharing and getting the most out of the data is absolutely critical.”
Improving safety and equity while protecting privacy
Mobility data-sharing isn’t limited to cars. The city of Detroit recently launched a partnership to pilot analysis tools for mobility data on dockless scooters and bicycles. The pilot will use data provided by scooter rental companies Bird and Lime.
The project is led by SharedStreets, a nonprofit that builds tools and data governance models designed to make it easier for cities and companies to work together to improve urban mobility.
“Essentially, maps inside different companies and cities all speak a different language and reference (data points) in a way that differs from each other slightly,” says Shared Streets co-director Mollie Pelon McArdle. What SharedStreets can do is “provide a common global reference for the street,” she says, so data collected from various companies can work together.
She notes the information is aggregated to protect privacy.
“We wouldn’t be able to see a single trip,” she says. “Personal privacy is preserved while aggregated data is still available to make informed decisions. Individual trip information is not necessary for transportation planning.”
Justin Snowden, smart mobility strategist with the Detroit Mayor’s Office, says the city was already collecting some scooter data on an “operator by operator” basis, but there “wasn’t anything that gave us a true aggregate picture.”
The partnership with SharedStreets will allow a comparative analysis across operators, and get an idea of how and where scooters are being used in the city independent of what operator is providing the scooter.
“One of the things we’re most excited about is using that data to promote equity in terms of distribution of scooters throughout the city,” Snowden says. “We know there is a huge concentration, across operators, of scooters downtown and in Midtown, the central business district. But this mobility tool needs to be available to residents across the city.”
Snowden says the data-collecting project will allow the city to better understand the distribution of scooters around the city and leverage that to encourage providers to place scooters in neighborhoods outside that central business district.
Focusing on efficiency
Mcity, the University of Michigan’s advanced mobility research facility in Ann Arbor, is also sharing its data, with an emphasis on efficiency.
A recent Data-Collection and Management-System project collected and archived data on Mcity infrastructure, including the traffic signals, vehicle detectors, and roadside units at its test facility. The data can be used for both real-time and offline analysis and visualization to support the testing of autonomous and connected vehicles.
A connected vehicle-based control interface device (CVCID) collected data from equipment at each intersection, and data was forwarded to the Michigan Traffic Center server database in real time. Mcity’s data is being shared with all Mcity Leadership Circle members and the city of Ann Arbor.
Yiheng Feng, co-principal investigator for the project, says urban planning efforts to address traffic congestion traditionally draw data from fixed-location sensors. The end goal for the Mcity project involves combining data from stationary infrastructural detectors with information from the city of Ann Arbor’s fleet of connected vehicles.
“(Traditionally), only if a vehicle passes the detector do we know where the vehicle is. Between sensors, we don’t know anything,” Feng says. “Now if we have connected vehicles in communication range, you get the full trajectory of the vehicles. That gives you spatial and temporal continuous data and provides much (richer) data than you can get from fixed-location sensors.”
Feng says the team doesn’t yet know all the potential uses for the data. His group is mainly focused on traffic applications, with an emphasis on efficiency, though Feng says the data could also be used for safety or sustainability applications as well.
A Troy-based company very recently launched the largest mobility data-sharing project in the country. nuScenes by Aptiv is an open-source autonomous vehicle data set offering 1,000 driving “scenes” created by collecting more than 1.4 million images and 390,000 Light Detection and Ranging (LiDAR) sweeps in Boston and Singapore, with more than a million 3D “bounding boxes” annotated by human experts. They represent some of the most complex urban driving scenarios available for study.
“The genesis of nuScenes by Aptiv came from a gap Aptiv identified in sharing critical data with
researchers and the academic community to enable advancements in (autonomous vehicle) research and safety,” says Karl Iagnemma, president of Aptiv Autonomous Mobility. “Aptiv believes that global knowledge sharing — between corporations, startups, academia, governmental agencies, and other entities — will enable robust progress and innovation in our industry.”
nuScenes represents the largest multimodal, three-dimensional autonomous vehicle data set ever released and shared with the public. Though the data was only released in March, over 200 academic institutions and over 1,000 individual users have registered to access the dataset, which Aptiv hopes will be used to improve the safety of autonomous vehicle technology.
“We encourage collaboration across the industry, especially within the research community,” Iagnemma says. “We have carefully published data that will lead to the development of critical safety standards in our industry.”
Sarah Rigg is a freelance writer and editor in Ypsilanti Township and the project manager of On the Ground Ypsilanti. She joined Concentrate as a news writer in early 2017 and is an occasional contributor to other Issue Media Group publications. You may reach her at email@example.com.
Steve Remias and Justin Snowden photos by Steve Koss. NuScenes screenshots courtesy of Aptiv.