People Moving | Data Analytics and City Mobility

It’s been fascinating watching Amazon preparing Crystal City – ahem, National Landing – for it’s new HQ2 complex. Promising a workspace for 25,000 people, almost as many as work within the Pentagon itself, the company has a lot of work to do.

And no, not just building a new building. Because when you bring 25,000 people into a city, you need to design a total ecosystem for them. Where will they live? How will they get to work? What city infrastructure has to be added to house and support 25,000 additional humans?

Transportation, for one! Crystal City is the focus of a number of different transportation investment programs, including a fully remodeled multimodal transit hub designed to potentially link the Maryland MARC, the Metro, the Virginia Railway Express (VRE), and the bus system. However, a lot of these projects are products of an old vision – the VRE station in Crystal City, for example, has a platform that’s only about half the length of the current trains and has needed to be expanded for a while – and may not be sufficient to handle the current commuting situation, let alone waves of new commuters to the tune of thousands.

DC already has the sixth worst traffic in the country (according to 2017 and 2018 reports) and while we have plenty of public transportation options, Deloitte’s 2019 City Mobility Index shows that 76% of commuters still choose to use their cars. Why? And what can we do?

From Deloitte 2019 City Mobility Index

Integration of Transportation Options

Considering that the DC metro area lies within the bounds of DC, Maryland, and Virginia, both in rural areas and municipalities, there are a lot of different transportation authorities to consider.

Within DC itself, WMATA has done a fairly good job of enabling commuters to pay for multiple transportation options – the Metro, MetroBus, and some local and commuter buses – with its SmarTrip card. The SmarTrip card is a small plastic fare card that you can purchase and load at any Metro station or online, reload conveniently, and swipe to get access on various forms of DC’s transportation.

However, you can’t pay to use the MARC and VRE with your SmarTrip card. So you’ve just effectively alienated everyone outside of the Metro limits – which, although they’re expanding out to Dulles International Airport and reach into both Maryland and Virginia, still don’t cover more than a fraction of DC commuters. As the city (and surrounding cities like Arlington, Crystal City, Alexandria, and others) expands, more and more people are commuting from the suburbs. And that always seems to add in at least a small amount of driving.

Expanding to the VRE, MARC, and more commuter buses enables your long distance commuters to not only get more versatility out of their fare purchases, but would allow WMATA to do something infinitely valuable – collect more data on travel habits.

Integration of Travel Data

It blows my mind that in this day and age, we are still relying to a large extent on people manually counting commuters and built in people counters on buses and other public transport. A much smarter solution exists, especially in DC, where many commuters are armed with their SmarTrip card, swiping in and out of stations and on and off transportation.

WMATA seeks to use “Trace,” a program that syncs SmarTrip card swipe data at station fare gates with the location of trains and buses to extrapolate how many people were riding that vehicle. That’s a start, but seems clunky when you think about all you could do with that data, especially when you start linking systems.

How many of your commuters are taking multiple methods of transportation to get to work? If a person takes a bus from Woodbridge to the Franconia-Springfield Metro Station, is that data captured? Is it used to synchronize timing of buses and trains (and the number of those available) during peak commuting times? It should be.

If that person parks at a Metro station, they will likely use their SmarTrip card to pay, so you can tell how often people pay to park at the station, and potentially, given billing addresses, where it might make sense to run a new bus route. Are we considering that facet? We should be.

Data collected over time, if SmarTrip were integrated into the VRE, would indicate how many of your riders coming into Crystal City from Woodbridge, Manassas, or Fredericksburg on the VRE then hop onto a bus or the Metro to get to Arlington or Pentagon City. We’re missing out on that data because the VRE doesn’t take the SmarTrip card.

We should be able to tell how many people take what form of transportation, what other methods of transportation that links to, and integrate our travel data into population profiles that enable us to sync timetables, offer efficient and linked transportation options, establish predictive maintenance patterns, and plan more efficient routes.

MIT has done a huge amount of work on integrating data analytics in urban transportation in their Department of Urban Studies and Planning . This overview is on my list of recommended reading.

DC Metro Transit app

Provide Transport Information Rapidly

It should tell you something if you’re a transportation authority that offers an app but yours is not one of the apps that gets recommended for use getting around town. The Metro app is clunky at best, the trip planner function is next to useless (seriously, whose time table defaults to 1900?), and while it does track locations of trains and buses and provide alerts, it’s nothing close to real time. And it doesn’t provide you data in return besides what people have searched for.

Think Waze. And Google Maps, who is trying really hard right now to catch up with Waze. Both are community-driven GPS navigation apps that allow other drivers (and walkers, bikers, and other forms of transportation in Google) to improve commute times by feeding real-time data to the app about travel times, road conditions, hazards, and other issues.

Waze also has an integrated travel plan function that lets you know when you should leave based on a preferred arrival time, or you can just use it immediately. They’ve also rolled out a Waze carpooling feature that lets you share gas cost and pay for rides by automatically charging the user.

So think about what you could do with that as a transportation authority.

First, you would need to incentivize users to share data. Waze does this by both helping you avoid traffic blocks and through gamification – the game elements the app adds to your experience, where you can compete for leaderboard spots, get thanked and appreciated for your responses, earn points, set your mood, change your icon…lots of little things that don’t amount to much outside the app but challenge people’s competitive urges and ensure sharing. Maybe you allow users to set your app to run in the background while linked to bus or Metro wifi and offer a number of free games to play with an easy option to switch to reporting.

The more data you have, the more you can provide users information about their transportation options. How fast or slow is their usual commute and are there other commute options available that might get them there faster? How does the express bus compare to the Metro that morning? This not only helps you get people where they’re going faster, but helps alleviate the strain on a transportation option that’s getting bogged down.

I’d be a lot more likely to pick up a Metro app if it let me plan my trips using a Waze-like interface, figure out the shortest or fastest commute option and see what other options there are to get me from my house to work, linked to Uber or Lyft so I don’t have to worry about parking my car at the Metro and coming back by bus and then figuring out how to get it, rewarded me for use with points or discounted travel or potential in-app purchases, and, while we’re at it, let me just scan my phone bar code or QR code instead of having to keep track of a separate card.

Just a little food for thought if you ever find yourself in a station at the Metro and wondering just how this process could be made better with data.

Leave a Reply

%d bloggers like this: