I recently got to sit down with Dr. Regina Clewlow, CEO of Populus, to talk about one of the most critical, yet unappreciated, pieces to improving our transportation system: data.
Whether it’s high-level data such as the number of miles driven by cars in the United States each year or specific data such as the number of bicyclists passing through a single intersection each day, access to information already plays an important role in the design of our transportation system.
Data will also play a large role as we prepare for autonomous vehicles (AVs). And robust research and access to data related to self-driving cars is one of our seven principles for maximizing the benefits of these vehicles.
The ride hailing, bike-share, scooter-share world we are beginning to live in – let alone emergence of self-driving cars – presents huge opportunities and challenges for changing how we get around. Congestion, convenience, cost, speed, equity, and emissions associated with transportation are all subject to change – for better or worse – with the emergence of these new forms of mobility.
Regina is one of the most respected people I know working in the new mobility industry. She has worked from all angles to improve how we get around, as a researcher in the academic world to leading roles in the private sector, including her current venture as the co-founder and CEO of the mobility data platform company, Populus. Here’s an edited transcript of a conversation we recently had about data and new mobility.
Jimmy O’Dea (JO): Okay, let’s start with the big picture: new mobility companies collect a lot of data through the course of their operations. What kind of data are important to you as a researcher and how can sharing it help improve our transportation system for everyone?
Dr. Regina Clewlow (RC): The data we really focus on are centered around GPS traces of vehicles – whether they are bikes, scooters or cars – to help cities better manage street space, curb space, and parking. Collecting this data helps us understand where people are going so that cities and regions can plan around those decisions at a higher level.
On the safety side, when you have better information about how people are moving, you can better design streets, and you may find you can dedicate more space to safe bike infrastructure, which is a win-win for everyone.
There is also data that can help us achieve emission reduction goals by answering some key questions: to what extent are mobility services reducing vehicle ownership? Are people traveling more or less when they start adopting these shared mobility services? These decisions obviously have a huge impact on total transportation emissions.
JO: In a recent Forbes article, you point out that data sharing could help all parties. What are some of the benefits to companies?
RC: Sure, if cities can identify where hotspots are, they can design pick up and drop off zones, which would help ensure that these vehicles don’t disrupt the flow of traffic and make them safer for people getting in and out of them, as well as pedestrians and cyclists. This type of coordination hasn’t really happened in a scalable way, but it is a key opportunity that’s on the horizon.
JO: Okay, so better curb management could help companies get riders in vehicles, but some companies are still hesitant to get behind the idea of sharing data. Why is that?
RC: It’s not just companies concerned about the competitive intelligence aspect of revealing their business models or about the proprietary nature of the data. A lot of people are also concerned about the privacy of users.
A key challenge with GPS trace data, particularly with services like Uber and Lyft, is that many trips are going to people’s physical homes or physical work addresses. If someone were to get a hold of enough breadcrumbs, they could recreate trips and then attach other data sets to identify specific people.
JO: So how can cities etc. get access to the data they need without compromising riders’ privacy?
RC: Many experts are of the opinion that certain data should not be made publicly available in its raw form because it can compromise individual privacy. But there are many ways to aggregate data so that certain elements are made publicly available without compromising personally identifiable information. Of course, too much aggregation should also be avoided because aggregation of data can start to render it useless for transportation planning and policy.
I really believe that data needs to be made available for researchers at national labs and at research universities in order to help us understand what’s going on and what the future looks like. This can be achieved without compromising privacy or proprietary information, which is precisely what we do at Populus for cities from coast to coast.
JO: Are there any examples that come to mind in other sectors of sharing data for the public good without compromising privacy?
RC: There are numerous examples, but one that I’m quite familiar with because I was an aviation researcher previously, is that in exchange for utilizing publicly-funded airspace and airports, commercial airlines are required to report on a 10 percent sample of all trips, including origins, destinations, stopovers, and fares. So, the FAA knows exactly how many people are getting on and off planes, and they also know with the 10 percent ticket sample how much people are paying – so what are the average fares for specific routes. All this data is made publicly available without compromising any personal information or business information, because there is a time delay in when the data is publicly released.
JO: Have there been any data sharing requirements for ride hailing services?
RC: Mobility services have rolled out very quickly, so even though there’s a clear need for data for transportation planning, at a high level, there is very limited data available to the public sector.
Cities have been frustrated by the lack of data that’s been made available to them by ride hailing services; for the most part, they have virtually no information. A key challenge that has emerged in the increasing privatization of mobility services is that these services don’t necessarily strive to meet public goals. The important, continuing role of the public sector will be to define policies that can help us meet goals such as improving safety, ensuring equitable access to transportation services, and improving efficiency – even as transportation services continue to become more privatized.
We’re starting to see some progress with cities requiring data from transportation network companies. Some examples are New York, one of the first cities that required trip data from Uber and Lyft, and DC, which just followed suit through their for-hire vehicles program.
With dockless bikes and scooters, cities have significantly more regulatory authority. There are a couple main reasons for this. One, with micro-mobility, bikes and scooters are small vehicles – cities can throw them on the back of a truck and impound them. Two, because users come and check them out and use them and then leave them somewhere, they’re stationary for a certain portion of time, and again, it’s easy to impound them. Three, they tend to be owned by mobility operators, whereas with ride hailing services, the vehicles were constantly moving, not owned by the companies, and the people driving them are technically not employees.
JO: I’m not aware of any massive data collection or GPS tracking of the trillions of miles being driven by personal cars each year. What makes data collection from new mobility companies so different?
RC: Actually, there are ways that cities can access personal vehicle data today. There are companies that aggregate connected vehicle data and sell it. Cities are making use of that kind of data. They actually just used it in a recent study on traffic caused by ride hail companies in San Francisco. But from a regulatory perspective, it’s a lot easier to establish policies that effect a few companies than trying to affect millions of individual drivers.
JO: Okay, I’ll end on autonomous vehicles. Most of the data we’ve been talking about so far is from ride hailing services like Uber or Lyft, or car, bike, and scooter sharing services. But autonomous vehicles are on the horizon and could be a major part of our transportation system. Should there be data sharing requirements for autonomous vehicles?
RC: A lot of cities are thinking about how to deal with the regulatory environment for scooters, bikes, and ride hailing companies and what that means for the potential arrival of autonomous vehicles. If AVs are rolled out in a mobility-as-a-service fashion, establishing data policies for ride hailing services could help pave the way. In addition, establishing the technical infrastructure to make use of that data and to monitor and manage mobility systems is something cities are thinking about now.
I believe it is fair for cities to require data sharing from private operators in exchange for the use of public right of way. Similar to airspace, if private companies want to use publicly-funded space, it is completely reasonable for them to pay for the utilization of that space, with data or dollars. In fact, many experts would agree that the appropriate pricing of physical space utilized by transportation services is one of the most efficient ways we can reduce traffic congestion, and ultimately the energy and emissions impacts of the transportation sector.
Dr. Regina Clewlow is the CEO and Co-Founder of Populus, a data platform for private mobility operators and cities to deliver safe, equitable, efficient streets. She is a former transportation scientist from UC Berkeley, Stanford and UC Davis, a former Clean Vehicles Kendall Fellow at UCS, and has been a leading expert on shared mobility and autonomous vehicles.