“Combine a ton of data with some analysis for grand results”—that’s the promise of big data, according to Jonas P. DeMuro in a December 2019 article in TechRadar.com. In the transportation realm, the ability to provide context to travel data seems like one of the greater opportunities of big data and analytics. However, big data can be nebulous, and our members have numerous questions as summarized in the article on
ITE’s Big Data Listening Sessions on page 24.
To illustrate the concept, my philosophy for workout tracking and contribution to a large pool of activity data is: if it’s not on Strava, it didn’t happen. Strava is one of many GPS-based exercise trackers that maps and analyzes activities. In aggregate, Strava and similar systems provide data on where and when walkers, runners, and bikers access the transportation system. The beauty of this data resource is that it is continuous and touch-free. There are no sensors to maintain, it is updated consistently, it covers a significant area, and it has a large sample size.
But beyond spatial and temporal data, other insights are conceivable, including: the age of active neighbors, purchases associated with activity, adverse weather impacts on activity, arterial crossing frequency, and property values correlated to activity. In a December 2020 article for Tech Republic, Mary Shacklett writes, “One of big data’s soft spots is the inability to identify impactful business cases that build revenue, reduce costs, and improve operations.” Herein potential exists. As the system experts, transportation professionals can work with data specialists to refine business cases.
A few cautions are necessary, though. Computers and machine-derived outcomes are not inherently neutral. As human creations, they are wholly subject to human error and bias. Further, algorithms are built on existing data and predict trends accordingly. If there is a flaw in the raw data, an under-represented stakeholder for instance, the algorithm is likely to perpetuate it. In an October 2011 article in the MIT Technology Review, Erica Naone notes, “About 40 percent of Twitter’s active users sign in to listen, not to post, which… suggests that posts could come from a certain type of person, rather than a random sample.” Similarly, in my Strava example, I self-selected to provide data to the system as a function of my need for data. Not everyone cares to do so. Further, I have leisure time, live in a neighborhood that is conducive to outdoor recreation, and have the means to purchase gym memberships. These factors limit extrapolation of Strava data to the greater population.
Big data also presents resource allocation challenges. Typical transportation agencies may not have the technology or talent to dive into the data lake. “It is complicated and mostly to access usable information fast enough to make a difference,” Jamie Carter pens in a January 2016 article for TechRadar.com, and experimentation requires iteration. And while people are willing to offer information to private companies, there is a distrust of data in the hands of government. This offers opportunities to partner with universities, expand student chapters, create public-privatepartnerships, and develop interest in the transportation field within other sectors. Big data opportunities abound. With consideration and caution, ITE professionals can enrich the resources used to model our communities.
This is the president's message from the February 2021 issue of ITE Journal.