The Oregon Department of Transportation (ODOT) has set a precedent for public agencies across the country with the purchase of a Strava dataset in order to inform their research. In the fall of 2013, the ODOT bought a one-year license of a dataset that includes over 400,000 individual bicycle trips made by about 17,700 riders covering 5 million miles logged on Strava that same year. The agency hopes that this data will serve to inform policy and project decisions.
The idea to use data from Strava, the popular website and smartphone app where people can track their bike rides, came to ODOT Active Transportation Policy Lead Margi Bradway one day while she was riding her bike. Bradway noticed other cyclists checking Strava and wondered if her own agency could use that data.
The problem, however, was that Strava had never given their data to a public agency before, and the ODOT had never purchased data from a private company. The biggest challenge turned out to be how to package that data in a way that both parties were comfortable with. It took several months of negotiations to strike a deal.
Still, with that purchase, the ODOT is the first organization to build a data feed that could transform how it makes decisions about their projects, plans, and policies. Pushing the boundaries of bicycling planning is an important step for the ODOT which is facing an unsteady future and wanting to shed its reputation of being out-of-date and highway-first.
Even though the ODOT only manages main arteries and rural roads, not necessarily city streets, the data set from Strava could go a long way to inform research. The biggest problem for many transportation agencies is the lack of data, and this gap makes it harder to justify infrastructure investments for bicyclists and plan for future bicycle traffic growth. Many cities do bicycle counts already but they only measure one location for a small period of time.
The data from Strava and other popular GPS tracking devices and smartphone apps could help to solve these problems. The ODOT is most excited about how this data set from Strava could provide context about how people ride. In other words, it could inform decisions makers about why bicyclists do what they do. The ODOT noticed that their bike counters were not capturing all the traffic on a bridge on Highway 101 because people choose a variety of ways to cross the bridge. In addition, bicyclists often behave like pedestrians in certain situations and take shortcuts in other places.
The ODOT began with formatting the raw data into something that their staff could access and use in their daily work. The agency has a special GIS Unit that provides geographic information products including maps and their transGIS tool. The ODOT can also use this data set for internal outreach and education for various management teams and different departments within the agency. There are many other possibilities where this data could be prove useful including travel demand models and forecasts, construction detour plans, and maintenance scheduling.
There is the problem, however, that the data from Strava might not be representative of all bicycle users. Most Strava users are those on serious training rides. Plus, the utilization of this data is still in its early stages. There is much to be learned about how this information could advise policymakers on how best to make decisions.
Even though the sample size problem exists, there is the potential for other research opportunities. The ODOT has already funded a new research project at Portland State University through the Oregon Transportation Research Education Consortium. The research is entitled Crowdsourcing as a Data Collection Method for Bicycle Performance Measures and assesses how smartphone technology relates to information gathering on bikes.
Other states have gotten wind of the ODOT’s purchase and hope to implement their own projects. The executive director of the Bike-Alliance of New Hampshire wants to use similar sample sets to update statewide bicycle maps, enhance their bike counts, and inform their discussions on where to place rumble strips.
More importantly, the ODOT’s purchase is a big step for public/private partnerships. Ultimately, it could have bigger implications for more than just bicycling planning.
“ODOT Embarks on ‘Big Data’ Project with Purchase of Strava Dataset.” http://bikeportland.org/2014/05/01/odot-embarks-on-big-data-project-with-purchase-of-strava-dataset-105375