Laura Schewel, CEO of StreetLight Data

As information technology continues to penetrate traditional industry sectors like transportation, the innovations just keep on rolling in. The latest case in point comes from a San Francisco-based start-up, StreetLight Data, co-founded by Richmond native Laura Schewel.

StreetLight Data combines a variety of data sources — wireless data, GPS navigation data, Census data — to infer mobility patterns in the United States. Providing a more detailed and nuanced understanding of where people travel, Schewel believes, will help retailers make more intelligent decisions about where to locate and will enable planners to make better decisions about where to invest transportation infrastructure dollars.

The company is new — founded in 2011, it announced its first round of venture funding this April — but it has some success stories already. Schewel cites how StreetLight data was used in an urban redevelopment project in Oakland. A group was trying to get retailers and restaurateurs to move into a depressed part of the city without much luck. The demographics of the people who lived there were unattractive from a business viewpoint. But StreetLight’s analysis showed that an affluent population that worked downtown and enjoyed a hip nightlife drove through the area every day. “Suddenly,” she says, “they had an attractive location to sell as opposed to a dangerous location.”

StreetLight Data’s main selling proposition is to retailers. Traditionally, retailers have relied upon demographic data regarding where customers live to inform decisions about where to locate stores. Now, thanks to StreetLight Data, they can base location decisions on the basis of the routes people travel. The data also can be used to target online advertising to the workplaces of people who pass by their stores on the way home from work.

A socially beneficial outcome would be if retailers placed stores in locations that enabled people to drive less, says Schewel, whose passion for transportation analytics grew out of a concern about climate change and a desire to reduce automobile C02 emissions. “If you drive the same drive home from work every day, it’s better for you if a grocery store opens up on your route,” she says. “You can chain your trips together. You can reduce extra trips.”

Transportation accounts for 70% of the nation’s oil consumption and one third of its climate emissions. Shopping-related driving accounts for 20% of all vehicle miles driven. In theory, aligning the incentives of drivers and stores will help change behavior, she says. “The American marketing system is the most powerful force of behavior change in the world.”

The company also has applications for urban planning and transportation engineering to better align transportation investments with how people use their streets and roads. State departments of transportation make decisions in a fog. They base investment decisions on traffic counts at specific locations and survey data. “They miss nuances,” says Schewel. “They really don’t know what happens after they institute a new policy. They can’t measure the before and after very clearly. We can.”

StreeetLight Data is crunching numbers that will result in a series of white papers later this year. Two of them will be set in Virginia. Unlike many state DOTs, whose data is inaccessible, the Virginia Department of Transportation puts it online and makes it easy to download. Thanks to that transparency, StreetLight Data is doing a lot of its validation on new data sets here in Virginia. “The state data is so good.”

Bacon’s bottom line: Roads are getting smarter — more sensors, more cameras, more data. Now our analytics are getting smarter. Perhaps the most remarkable thing about StreetLight Data is that it provides a feedback loop to connect transportation decisions with land use decisions. We are on the verge of a remarkable transformation.

Share this article


(comments below)


(comments below)


14 responses to “Smarter Analytics for Smarter Roads”

  1. larryg Avatar

    things are going to move very fast on this. Not only will we be able to collect data for different geographies and demographics but we will be able to COMPARE!

  2. reed fawell III Avatar
    reed fawell III

    Streetlight Data is producing data that is potentially very useful.

    Question – will those tasked with shifting through, selecting out and interpreting that data be up to the task of delivering a useful product to those who will put it to its best and highest productive use?

    Often unfocused volume and minutia of information defeats it application.

    1. Neil Haner Avatar
      Neil Haner

      Reed – I think the key to making sure the data turns into a useful product is by leaving it in the private sector’s hands, as the example of Miss Schewel above demonstrates. Leave it up to VDOT or localities and it becomes just another bureaucratic dumping ground. But private corporations have profitability driving them to find quality applications for the data.

      The counter-argument to leaving it in the private sector’s hands, of course, is “privacy.” People with concerns of a Big Brother watchdog state or at least identity theft are going to point to this as another window for their boogeymen to put eyes on their personal information and whereabouts… but I’m pretty sure that in this day and age, there are plenty of avenues for that even without this particular technology on both the figurative and literal streets

      1. Although I didn’t mention in the blog post, Schewel made a point of stressing the privacy protections built into their business model. StreetLight Data does not get IDs for individual cars.

        1. larryg Avatar

          unless of course the FBI show up and want someone ‘tracked” eh?

          I note GOOGLE and others have tried in vain to “anonymize” data only to get pressure …..

          so I do not think Schewel ought to stress that idea too much and here are two reasons why:

          1. – look up “stingray” in Wiki and then answer that such a thing could not be used in gathering data?

          2. – the very same technology used in tolling now days – license plate “readers” can and is deployed in a variety of places these days besides toll booths.

          besides, one or two more Boston incidents and people are going to demand more “security”…

          and there is one very important analytic that would be useful and that is known as origin-destination studies.

          they don’t do them much anymore because stopping traffic to take polls is not tolerated like it used to be.

          but in the old days, you’d set up a traffic check and ask where someone started their trip and where they were going to end up and the aggregate data could be very useful in determining what infrastructure was needed where and what size, etc. That info could be captured from reading plates… and building aggregate data profiles.

      2. reed fawell III Avatar
        reed fawell III

        Yes, I agree generally with Neil. And I would think the privacy issue can be rather easily overcome here, as the CEO suggests.

        Still the real genius here will be in focusing the raw data for delivery in ways that assure its pertinence and simplify its interpretation because the potential advantage gained from the story it tells is huge if it’s delivered to people able to see its significance. It’s remarkable how few people do.

        1. larryg Avatar

          well the private sector is not going to fail to understand the significance nor the importance of the data if they are going to risk money toward some profitable end and depend on it. One can count on that.

          That’s not necessarily the case with those who would use public money toward some end and that of course includes those who would try to convince the govt that a private development had merit.

          I like the contrast the difference between a VDOT traffic study and an “investor grade” traffic study. The latter is risking money based on the traffic so they are a lot more concerned with the fidelity and interpretation of the data than someone trying to convince that a new road should be built with public funds.

          1. reed fawell III Avatar
            reed fawell III

            I wish that were true, Larry. Far too often it is not.

          2. larryg Avatar

            let me revise – MOST private sector endeavors involve putting money at risk that could be lost and if it belongs to an individual that generally is not a good thing.

            when money is at risk that belongs to the generic “public” there is far less care and concern about it being “lost”.

            any better?

          3. reed fawell III Avatar
            reed fawell III

            A bit but often times the private active party and all his accessories simple passes off all the real risk to public investors and secures himself and his accessories with upfront fees and operating charges thereafter.

            This is why the sub-prime securitization example is pertinent. And, I suspect, it is also quite typical of public/ private road deals as well.

        2. larryg Avatar

          re: passing risk on to others aka securitization/sub-prime and now probably PPP.

          well… in the case of the beltway HOT Lanes (as an example), the state still owns the infrastructure and Transurban has a contract and a lease.

          If Transurban goes belly up – it appears to me the state is protected but I realize there are probably some not obvious “gotchas” but that would be true in virtually all private business relationships with the govt.

          I basically support the govt having the private sector doing as much work as possible and to keep employee overhead to only what is required and cannot be done by the private sector.

          Of course that means the govt has to be “smart” and have “smart” people or willing to employ “smart” consultants and even then screwups will happen.

          but that’s different than nefarious skullduggery…. on one or both sides and that’s endemic to ……… mankind…. with govt as one of the players,

      3. larryg Avatar

        I totally agree with Neil. some things are “public goods” that VDOT takes care of but other things have tremendous marketplace value and most of the time, most govt organizations are not interested and in fact, the private sector will complain if the govt tries to engage in entrepreneurial endeavors especially if they are “too” profitable.

        FDIC and Va. Inland Game and Fisheries make a point about not costing the taxpayers a dime but let them try to improve on that and make a profit to return money to the Treasury – and the private sector will holler like stuck pigs.

        The state parks in Va are right now under pressure from private campground operators for selling campsites “too cheap”.

  3. reed fawell III Avatar
    reed fawell III

    This information would be valuable to local government land use and city planners, for many applications.

  4. […] Data aims to optimize retail locations and minimize travel time by utilizing data based on the routes people actually use, reported Bacon’s Rebellion. Co-founded in 2011 by transportation sustainability advocate […]

Leave a Reply