In a typical hour, more than a quadrillion bits of data – little pulses of zeros and ones – are transferred through the internet as jolts of electrical voltage and flashes of laser light across a massive global infrastructure of servers and routers to instantly serve your favorite cat video to your iPhone.
The tangle of this world-wide web is managed by clever software routines and algorithms that find the best ways to send information from place to place – say, from Amazon’s data farm in Oregon to your smartphone in Arlington via a series of internet hops. It is this software that makes the web act like clockwork.
Advances in such real-time routing software is on the verge of making transportation networks smarter too. People and goods are constantly moving in and around cities using roads, rails, trails, bridges, and waterways, just like bits of data are buzzing through underground cables and cell towers. Data about where travelers and shipments are coming from and going to can be input into logistics algorithms to make the process more efficient. For example, UPS and FedEx are constantly improving their performance by tweaking their pick-ups, deliveries, and routes. And in Washington, D.C., alone, companies like Uber and Lyft and Split (my former employer) and Via and Bridj have tested their first generations of demand-responsive transportation solutions on the streets.
In practice, these private enterprises are creating a new software layer that seeks to serve up efficient transit at a lower cost. These “microtransit” businesses are seeking to upgrade how millions of trips taken during rush hours each morning and evening, using smart software to efficiently route groups of riders through a web of roads.
For the internet, however, a large portion of essential software is open source. Rather than companies like Google and Microsoft selling propriety packages, thousands of hours and millions of lines of code are given by a community of developers who believe that internet software should be free and open to all.
For example, the free and open-source Apache HTTP Server is the world’s most-used web server software, used by everyone from Fortune 500 companies, to Mobility Lab, to me and my wife on our travel blog. Having the underlying code free for all to read and edit makes the internet more secure and faster to evolve. Errors and bugs are quickly found and squashed, and new features are added as they’re dreamed up.
The open-source model is the right path for city transportation software too, rather than having companies create proprietary platforms and compete to win city contracts. Innovative cities already are looking to add software-based, smartly-routed shuttles to their menu of transit options. So far, they’re largely turning to the private sector: Helsinki, Finland, built its Kutsuplus program with Ajelo (which was acquired by Split); Kansas City is piloting a service with Bridj and Ford; Orange County is working with Via and Mercedes; and there have been solicitations for bidders to launch their software platforms in Santa Clara, Tampa, and Baltimore. There’s also TransLoc’s OnDemand software being deployed by universities from Emory to NYU, and Ford’s Smart Mobility program being piloted on their campus in Dearborn.
Helsinki’s former Kutsuplus microtransit program
What if the first step to launching a new microtransit service in, say, Arlington County, Virginia, were as easy as downloading an open-source software suite to ART’s servers and mounting iPads in its minibuses, rather than a lengthy, convoluted “request for proposal” process? ART’s information technology team would download software from the open-source website GitHub, plug in publicly-available data about local travel patterns, customize logos and text, and hand over a complete, functioning on-demand software system to fleet managers and to residents in the form smartphone apps and websites. Behind the scenes, open-source coders and engineers would keep the whole thing running smoothly. Later on, ART could deeply customize its platform and contribute to the software package for other cities, towns, and counties across the globe. A community of innovative cities would become participants and contributors, completing the positive feedback loop of development and deployment.
With this in mind, the U.S. Department of Transportation should help build and test open-source, demand-responsive transit software. Working with a city or group of cities, DOT should fund the initial teams of operations researchers who design algorithms, software designers who code the back-ends and front-ends, and city transit agencies who run on-the-ground pilot tests. Such a program would be a mash-up of Kaggle’s taxicab routing algorithm competition and DOT’s own Smart City Challenge, in which next-generation public software is driven on public streets by public agencies. Similar to the move to “open data,” we need “open algorithms” to put cities and their citizens back in the drivers’ seats of public transportation.
Cities have always centered around the people who live, work, and play in them. So-called “smart cities” are all about software: using smartphones and supercomputers to optimize sustainable energy, transportation, and behaviors. Let’s take steps to make sure that citizens, not companies, maintain ownership of our cities.
Photo, top: “Traffic Light Tree” sculpture in London (Gaellery, Flickr, Creative Commons).