On paper, the task was gargantuan. To slow the rapid spread of the coronavirus, the New York City subway would start closing every night for the first time in 115 years. That meant the Metropolitan Transportation Authority (MTA), its overseer, had to create a massive bus network to mirror 665 miles of track. That’s roughly equivalent to a line stretching from New York City to Chicago, for a system that was serving around 5.5 million people each weekday. And they had to do it fast.
Shortly after the decision was made, at the end of April 2020, agency planners logged on to Remix, one of the most popular transportation planning platforms in the world. The self-service software allows transit agencies to easily reroute lines or create new ones. The MTA had already used it to begin redesigning the Queens bus network.
But for this new undertaking, the agency would need more data. The MTA wanted to know: Where did the frontline workers who rely on overnight service live? And where were they most likely headed? With that information, its planners could plot the most efficient bus routes and maybe even serve these commuters better than the subway ever did.
At Remix’s headquarters in San Francisco, a team of software engineers sprang into action. The Remix team pulled data from a variety of sources and plugged it into the tool to help find the best routes. As a result of this work, the MTA added three new bus routes that would connect, for example, health-care workers and others in the Bronx and in south and east Brooklyn to the west side of Manhattan or any point along the way.
Days blurred into nights as work continued on both coasts. Ultimately, a task that would typically have taken weeks, if not months, was finished in a few days. On the evening of May 6, New York City’s subways shut down, and the new night bus network flickered on.
It was an early battle in the worst existential crisis for Western urban public transportation in our lifetimes. Across Europe and North America, ridership has plunged amid pandemic travel restrictions. The traditional hub-and-spoke model, with transit networks designed to flush people in and out of a central business district, was upended. Rush hour, as we know it, suddenly became less of a rush.
Tiffany Chu, Remix’s cofounder and newly minted CEO, watched this all play out in real time. “Transit agencies are always changing things up, but most of [those changes] are small, incremental ones,” Chu says. “When covid hit, immediately we saw in Remix’s admin panel just a lot of rapid changes happening: 50% service, then 30% … agencies were just slashing service everywhere.”
A year later, this unprecedented shock to modern mobility is still reverberating. The long-term shift to remote white-collar work is casting doubt on whether rush hour will ever fully return. The office is in retrograde. And for transit systems, the implications are profound.
The pandemic, Chu argues, poses a fundamental threat to transit agencies in the West. “Agencies are just being forced to learn how to do more with less,” she says. But she believes this massive system-wide disruption is also a rare opportunity to rethink public transportation for the better.
Opportunities over patterns
When Chu, now 32, graduated from MIT in 2010, she looked to pair her skills in the then-burgeoning field of user experience design, or UX, with her interest in cities. That took her to gigs at major architectural firms, a stint writing for the design publication Dwell, and a position at Zipcar, where she was the car-share company’s first UX hire. But she felt pulled to do more.
Friends told her about a fellowship at Code for America, in San Francisco, a year-long role focused on technology aimed at making government work better for people. She applied and soon packed her bags for California. That’s where she found some like minds: designer Sam Hashemi and engineers Dan Getelman and Danny Whalen.
“We were all vaguely interested in transportation,” says Chu from her bright San Francisco apartment, a bike in full view on Zoom. The field held a particular appeal: thanks to the transportation sector’s long-standing commitment to open data standards, she explains, “there’s a lot of programmatic data you can work with that doesn’t exist in other realms of civic tech.”
The quartet worked on a hackathon project, a typical techie icebreaker, a few months into the fellowship. After hearing friends complain of bus routes that never seemed to match up with where they wanted to go, they devised a widget that allowed users to suggest new routes to San Francisco’s transit agency. They named their tool Transitmix, inspired by Streetmix, another hackathon project that has since become a popular street design platform.
When their fellowships ended, Hashemi rallied the group to pursue a product geared toward cities. Code for America seed funding and an investment from Y Combinator led to Remix, which became one of the accelerator’s first gov-tech startups. Hashemi was the CEO; Chu was the chief operating officer.
Remix’s primary software appears, at first, to be a kind of cost-benefit calculator. When a planner plots a new route across a map on the screen, the platform estimates how much it might cost and who might ride it in view of who is able to access it, ultimately helping planners assess whether it’s a worthwhile public investment.
Adding more data deepens the tool’s technical analysis: with a few clicks, demographic information and existing ridership figures help planners visualize the routes that would best meet a community’s needs. There’s also a Google Docs–like element: users can leave notes for one another suggesting that, for example, two stops should be consolidated into one. The platform’s design makes it easy to share maps and routes that are fluid, clear, and intuitive—in a planning process that, for so long, has been none of those things.
“There’s a widespread perception that urban transportation planning is somehow a technically complex problem when, in reality, it’s a politically fraught issue with long-standing, well-understood solutions.”
Jarrett Walker, a renowned transit consultant, was one of Remix’s early advisors. In addition to cost and service levels, he suggested adding travel time—how long would it take people to get around, and what choices would they encounter along the way? So the group built what has become one of the platform’s most popular scenario planning tools: “Jane,” a rider isochrone, or travel time indicator, that shows all the places she could reach in 15 minutes, 30 minutes, and so on.
That advice jibes with one of Walker’s guiding philosophies: that public transportation planners must focus on fostering opportunity rather than predicting patterns. A similar sentiment is now gaining traction in policymaking circles: access, not ridership, should be the measure of success. And the right technology can help cities deliver on that promise.
But agencies’ capabilities vary immensely, says Evan Landman, a transit analyst at Walker’s firm, Jarrett Walker + Associates. Some are highly sophisticated; others are still using tab-laden Excel spreadsheets. Plotting a new bus route can take forever. Whenever two agencies have to collaborate and bridge that divide, the pace drags.
No doubt there are problems that tech cannot solve—repairing institutional trust, for one—but often, agencies just need help exploring the options, Landman argues. “There’s a widespread perception that urban transportation planning is somehow a technically complex problem when, in reality, it’s a politically fraught issue with long-standing, well-understood solutions,” he says.
Remix doesn’t try to solve the politics with a “magic cure,” Landman adds, but strives to show how decisions may affect individual riders and their landscape. “It’s really useful in helping to explain the different sides of a question that, ultimately, doesn’t have a single technical answer,” he says.
A complicated relationship
Shortly after Remix’s launch, in 2014, a transit manager at Oregon’s Department of Transportation reached out: rural agencies in his state needed better tools. He became their first customer. “We were over the moon and just bewildered that someone would pay us!” says Chu.
Then came Bay Area agencies, then Miami-Dade, then Chicago. Before long, Chu found herself Googling small Finnish cities as Remix signed up its first international clients.
Nearly seven years later, Remix now has a team of around 70, and a client list that includes over 350 transit agencies across five continents, including titans like the MTA and Transport for London.
Every day, more than 240 million people worldwide interact with planning decisions made on the platform, from individual routes to system-wide overhauls. In March, the New York–based ride-sharing company Via acquired Remix for $100 million. (Remix will operate as a Via subsidiary, and the company says that Chu and the rest of the staff will stay on.)
Dan Getelman, Remix’s chief technology officer, says one of the team’s goals is to free up time for transit agencies to experiment more. “It’s always frustrating as a transit rider when you say ‘I guess this made sense at some point, but it doesn’t match the [riders’] needs or doesn’t feel reactive to what’s happening,’” he says.
The tech sector has a complicated relationship with public transit, though. On the one hand, technology has brought some urban infrastructure into the 21st century, easing passengers’ journeys with advances like software APIs (think subway countdown clocks), contactless payment, and navigation apps. But on the other, tech is a direct competitor; companies like Uber have been criticized for intentionally taking riders (and revenue) away from public transit, while simultaneously clogging streets. How the two can best coexist is an ongoing debate in both worlds.
Remix falls, perhaps, into a different category. It’s a tech company that goes all in on the public sector, betting that riders will be attracted to traditional public transport options with good, reliable service rather than an entirely new product. It’s a high-tech solution, sure, but the premise is shockingly low-tech: build it better, and they will come. And in our rapidly changing world of mobility, Getelman says, responsiveness is essential: “Being able to do that makes for a better system.”
A sort of transit inversion happened when covid-19 hit. Yes, city centers emptied, but ridership outside of central corridors—along local routes and at neighborhood stations—didn’t disappear entirely, and in some cases it actually increased. Riders were still moving; it’s just that where they were going had changed.
Local trips like these have typically been overlooked by planners making transit decisions. They involve fewer riders, and funding is tied to ridership. Race and class also play a role; poorer riders and people of color, who are more likely to live farther out and are less likely to own a car, have long been left out of citymaking.
As a result, the quality of these public transit trips deteriorates, which drives down ridership. Seeing fewer riders, agencies inevitably cut service, and ridership slides further still. This produces the transit version of a death spiral—a more arduous commute and fewer opportunities for the affected communities.
But since the pandemic began, Chu has seen a notable change in the data sets transit agencies are asking for. Instead of primarily requesting information about which jobs are located where, a question that has long shaped planning decisions, Remix is now helping cities evaluate how easy it is for residents to access essential services such as health care, education, and food.
The change is a welcome one. “If you look at just jobs, that is probably not going to tell you the full story,” Chu says. “You also need to look at very basic needs, like where you can get fresh produce in grocery stores. That is one of the most important metrics that isn’t usually discussed when you talk about transit accessibility.”
In October 2020, Chu wrote on Forbes.com that covid-19 was not “the death of the city,” as many critics proclaimed, but rather could foster the “rise of the neighborhood center.” With mobility restricted, people were forced to revisit what was close to them.
This reckoning has reignited interest in the “15-minute city,” where pedestrian-centric environments and responsive public transit put essential services within reach of a brisk walk or short bike ride. Transit agencies, Chu argues, should take note. “The 15-minute city isn’t supposed to just be where the buildings are tallest,” she says. “It should be based on a truly livable neighborhood center.”
Look close enough, and you can already see changes happening. The new tool kit of streetscape interventions that cities developed in response to covid-19 hints at a different urban future—one that includes more “slow streets,” which limit through traffic; pop-up or permanent cycle lanes; outdoor dining; and parklets.
“It’s investing in small downtown retail corridors, and people frequenting them more often during all times of the day,” says Chu. “You want a constant flow of all types of trips, to and from where people are gathering and where businesses are.”
Remix’s portfolio is following suit. A “streets” platform allows agencies to tinker with everything from car-free thoroughfares to expanded sidewalks. And a “shared mobility” tool for emerging “last mile” systems such as e-bikes and e-scooters lays out nearby destinations that can be made more accessible.
Over time, prioritizing the places people live instead of just the places they work will mean expanding the mission of urban transportation, Chu says. Planners should be able to more easily anticipate people’s needs and adjust accordingly—as the MTA did with its new pilot bus route from public housing sites to vaccination sites at colleges in central Queens and Brooklyn.
In a crisis that laid bare the immense imbalances in how we get around—who has access to what; who has to travel how long; and who, ultimately, has to put their lives at risk—it’s time to be more flexible, Chu argues. That means helping people get to where they’re going by creating services that are genuinely better—or simply making it easier for them to navigate their own neighborhood without a car. And it means starting to right the historical wrongs in our cities that have left so many urban denizens stranded.
“If you don’t change the rules,” Chu says, “nobody’s going to be able to change the outcome.”
John Surico is a journalist and urban planning researcher.
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
The Biggest Questions: What is death?
New neuroscience is challenging our understanding of the dying process—bringing opportunities for the living.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
How to fix the internet
If we want online discourse to improve, we need to move beyond the big platforms.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.