Convergence: How Can Better Data Improve Transportation?
At the Eno Center’s March 24, 2016 Convergence, one of the panel discussions explored ways in which better data (better collected, better shared, better analyzed) can improve the efficiency of transportation systems. The panel was moderated by the Chief Data Officer of the U.S. Department of Transportation, Dan Morgan, and the panelists were Shyam Kannan (Washington Metropolitan Transit Authority, WMATA), Prachi Vakharia (RideAmigos, a consulting firm), Joung Lee (American Association of State Highway and Transportation Officials), Wade Rosado (Urban-Insights, a consulting firm), and Chris Zeilinger (Community Transportation Association of America, an organization that promote public transportation).
At the outset, the moderator noted that while USDOT collects a significant amount of data, they do not use of all it productively, and despite having all of this data, it is not uncommon for USDOT to have a need for third-party data. Moving to the panelists’ opening statements, Lee mentioned the critical role of performance and data-based decision making, and how state DOTs need to use up-to-date data to make informed decisions.
Zeilinger opened his intervention by mentioning the role of data collection and whether it should be done just for its own sake or to improve the transportation system. According to him, we mostly have a “faith-based transportation” system and to improve on it, stakeholders must make sure that they are showing the value of data-based decisions to local authorities. During his opening statement, Kannan highlighted how traditional models still show a never-ending increase in driving, while real-world data is showing stagnant demand for driving and increasing demand for bicycle, pedestrian, and transit use. Vakharia stressed the need to use data to achieve specific goals, be it less emissions or different mode shares, for example. Finally, Rosado added to that idea, by noting that data can be used to balance mobility and leverage existing capacity by, for example, moving use to off-peak hours.
The conversation then moved on to the issue of access to data. Citing the example of WMATA, Kannan argued that many agencies are releasing data publicly, and large numbers of people are using that data, whether just for fun or to suggest improvements. WMATA welcomes all uses of the data it releases, he also added.
The other topic highlighted during the conversation was privacy. The moderator opened this topic by mentioning that users often don’t understand what data is being collected and shared, and perhaps agencies, academia, and industry should be using transportation research money on privacy protections. Lee suggested that perhaps a way to get users more involved in data and privacy matters was to create incentives, be they motivational (the way fitness trackers show their users their own data, for example) or even monetary.
Moving own to the Q&A session, the questions revolved around use of data by the federal modal administrations within USDOT, access to users personal data, and quality of data.
The first question directed at the panel was about the use of data by the different modal administrations within USDOT (such as in the NextGen program to modernize air traffic control at the Federal Aviation Administration and the Positive Train Control requirement for U.S. trains at the Federal Railroad Administration), and how these programs could be improved by having the right information, delivered to the right people at the right time. Zeillinger replied by mentioning that there is more data than we can use, and these programs need to balance technology with human factors, for example, regarding automation.
The next question was related to privacy and how users can access the data their use of a system is creating. Morgan mentioned that there might be a market opportunity there, in the sense that solutions can be created to help users make sense of their data. Finally, the panel was asked if good decisions could be made without making sure that the data is credible and good, and if the stakeholders are ensuring the quality of the data. The panel unanimously agreed that the answer to both questions was no.