Transit startups, playing nice with municipalities, machine learning and urban tech, and the enduring value of the bus
Ed. This piece was originally written on 14 November 2014. Bridj subsequently shut down operations in Boston, although the brand and concept was acquired and, at time of writing this intro, runs in Sydney, apparently without Matt George on board. Bridj was an early example of a service concept that is increasingly familiar, and by actively liaising with public transit agencies and municipalities, they provided a welcome alternative to the aggressive blitzscaling tactics of other mobility startups at that point.
As research for my recent Dezeen column, and the longer piece it was drawn from—both of which unpick the promise and pitfalls of predictive urban analytics and transport startups—I asked Matt George, CEO of Bridj, a Boston-based startup, a few questions. I’m reproducing the exchange below, as I find Bridj really interesting for a number of reasons.
Bridj clearly sits alongside existing public transport services, looking to augment and ‘bridge’ the gaps, and so potentially avoid some of the issues of the ‘adjacent incumbent’ I wrote about. Although they’re VC-backed, they focus on something akin to ‘public transport’ (albeit not run by a municipality), in the form of buses, rather than simply shifting private cars around (as per Uber, Lyft, Relayrider, Sidecar et al.)
They use big data, including from social media, combined with predictive analytics, which enables them to move the bus-stop to you, rather than you moving to the bus-stop. Yet they also work with existing routes and demand patterns, and have hired solid transit experience in the form of fomer Chicago and Washington DC transport head Gabe Klein.
Bridj also indicates the ongoing march of the technology industry into the urban sphere, or the future cities sector. These emerging companies are changing the ways cities work, in entirely new ways. And when did you last hear of a new municipal bus company started by a 23 year-old? Yet George is not quite the typical tech-sector brogrammer-CEO: he’s a biology graduate who took a year off to be a White House intern, where he worked with trend-predicting data. Perhaps as a result, Bridj are looking to collaborate with muncipalities and regulators—”unlike some other companies”, George points out—which is welcome indeed.
You never quite know which startups will go on to define new sectors, and it’s very early days in the field of ‘public transport startups’ and predictive analytics. None of us have a crystal ball to know whether they’ll succeed or not. And there’s lots I’d quibble with, even though I’ve never used it, not least the design of the buses themselves.
But for the reasons outlined above, I’d like to see Bridj find time and resources to continue to develop their services, and I hope more services like Bridj emerge. Many thanks to George for the replies below (which he freely admits he sent from the train, so do take that into account.)
Q&A with Matt George, CEO of Bridj
Dan Hill: How do Bridj’s predictive analytics work? (approximately) It uses social media patterns, but how does it pick up on patterns of those not using social media? Are you digging into social media to discover origin:destination information? Or hotspots of activity? Or all of these?
Matthew George: Our algorithms take into account about 19 different data streams including social media and MANY other sources to determine origin and destination information. Everything we have is anoynimized, so we’re looking at overall city movement patterns and not individual movement patterns.
DH: Will it stretch to areas of the city with low volume, as public transit has to? Traditionally, timetables, routes and universal services have been a way of stretching a service over an entire city. Can you explain how Bridj might do the same, through ditching routes and timetables etc?
MG: We don’t necessarily ditch routes and timetables. Especially during high predictability portions of the day, Bridj runs very similar to a intra-city commuter bus route with limited “responsive” pickups and dropoffs. For instance, some of the Bridj routes have defined departure times, while some run headway service during peak periods.
By combining demand response and scheduled service, the vehicle can adapt to geographic and temporal changes in transportation patterns.
DH: Equally, municipal transit agencies have to provide for those without phones, never mind no smartphones, social media presence etc. How will Bridj serve them (if at all)?
MG: Just to be clear, social media presence is an relatively small part of how our planning process works. Right now, our system is accessible by anyone with an internet connection. Users can make reservations online, and don’t need a smart phone (or any phone at all) to board the bus.
DH: How do you handle privacy, and anonymity—two key aspects of living in cities—and balance that with the possibility of seeing public patterns?
MG: None of the information we look at is personally identifiable. We do look at patterns of users who use our system, but that data is anonymized as well. At least with our own users, using their anonymous data makes their travel within the city easier and is thus a pretty big win for both parties.
DH: Do you liaise with public transit authorities, in terms of sharing data back and forth?
MG: We absolutely share data with public transit agencies all over the world. This is an incredibly (important) part of our collaborative effort to improve transit in cities
DH: Is your aspiration for Bridj that it sits alongside municipal transit services in cities, filling in the gaps? Or do you think it could replace such transit services?
MG: Bridj is a third option. On one end of the spectrum you have low cost low flexibility services like traditional mass transit. On the other end of the spectrum is high cost and high flexibility options like owning a car or using Uber. We are looking to be a third option-moderate flexibility and moderate price-that we think captures most of the needs of city travellers.
We’re discussing various scenarios with transit agencies all across the country and globe to supplement, but not replace, their existing service through direct partnership.
DH: Could you see that it is possible for municipal transit services to use similar techniques ie. predictive analytics, on-demand transit etc.
MG: Agencies are already using both predictive analytics and on-demand transit, however they are using it to analyze their existing service, or using it in edge cases (like super rural transit). What we offer is the ability to start with a clean slate.
(So) yes and no. We are an incredibly high-paced technology company who focuses exclusively on creating a better demand responsive mass transit system. There are some small agencies that we think can implement better demand response, however, most agencies are trying to simply maintain the service they have, and have little capacity to try new things.
DH: At the moment it looks like a costly service—but is it? How does it compare to monthly costs for public transit? Could they be comparable?
MG: During our beta period in each city that we launch, prices are $1 for super off peak, $2 for off peak, and $3 for peak service. Over time, we expect the average ticket price to be about $4-$5. On the cost scale, that places Bridj way closer to mass transit than it does to taking a taxi, an Uber, or owning your own car.
DH: How important is the experience of the ride itself? Your buses look relatively luxurious, compared to utilitarian public transit-is that part of the brand, and/or a core feature, or just what you think things should be like?
MG: We really think that EVERY part of the transit experience is important. The most important factor that our users love is reducing their travel times throughout the city by up to 60%. After that, they love the fact that they get a seat each and every time, and the cabin environment is comfortable with nice features like power plugs and WiFi. However, the on-board experience is not what we prioritize, rather we focus on developing more efficient transportation systems that get our users where they need to go for as little time, and for as little money as possible.
DH: Are there any local legislation issues you’re facing, as per Uber, Lyft et al?
MG: Unlike some other companies, we’ve taken a collaborative approach with our local municipalities. We push them a little more quickly than they’d like sometimes, but we have (or are in the process of!) all the regulator approvals we need.
DH: How scalable is your software? Could you boot up in another city relatively quickly? Do you plan to?
MG: We only build scalable software. It’s ready to be deployed to a second city now, and we have imminent plans of opening a second city.
Ed. This piece was first published at cityofsound.com on 14 November 2014. It provided supporting research for this piece, on transit and micromobility startups, predictive analytics and machine learning in governance, and urban planning and urban tech:
Leave a Reply