Swiftly Has Arrived

Today, we are excited to officially launch Swiftly — a more accurate, seamless, and community-driven urban transportation app.

At Swiftly, we believe the future of urban transit is multi-modal — a trend we already see. As city residents who avoid the costs and burdens of car ownership ourselves, we rely on a combination of mobility options — from walking and biking to public transit, carshare, rideshare, and even scootershare.

We are not the first to notice this multi-modal trend: there are already others that aim to integrate these options. They all, however, suffer from a shared fundamental issue: inaccurate real-time information. For over a year, we have been developing an algorithm to better predict real-time transit arrivals as well as a crowdsourcing platform to provide more accurate and timely advisory information to riders.

We are now ready to make Swiftly available to the public — starting in San Francisco.

MEET TRANSITIME

As we outlined in a prior blog post, the real-time public transit information that exists today in San Francisco — and, in fact, most major cities — could be much better. To solve this problem, we developed Transitime. Transitime is a next generation real-time transit prediction algorithm created by Michael Smith, the former CTO & GM of NextBus, and now a co-founding member of the Swiftly team. This algorithm leverages real-time and historical data inputs to better predict when transit vehicles will arrive.

To test Transitime's accuracy, we compared predicted arrival times against actual arrival times for every route and stop in the entire city of San Francisco — a couple hundred million transit arrival predictions. As it stands today, Transitime is nearly 20% more accurate than the official San Francisco system powered by NextBus — and we’re just getting started. We will continuously improve and refine Transitime.

  Prediction accuracy in the city of San Francisco. The x-axis represents how far away the vehicle is from the stop when an arrival prediction is made, and the y-axis represents how often the prediction was accurate at that distance. Swiftly defines prediction accuracy as described   here  .

Prediction accuracy in the city of San Francisco. The x-axis represents how far away the vehicle is from the stop when an arrival prediction is made, and the y-axis represents how often the prediction was accurate at that distance. Swiftly defines prediction accuracy as described here.

EMPOWERING THE COMMUNITY

Even the best prediction algorithms break down when real world transit issues — such as a collision or an equipment failure — occur. Getting advanced notice of these issues can make or break the daily commute. While official agency advisories are helpful, they are often delayed. To provide riders with more timely advisories, we built a crowdsourcing platform directly within Swiftly.

With over 45,000 riders in San Francisco currently using our public beta app “Swyft”, we can often detect and contextually notify riders of transit issues before official agency alerts. Take an example from San Francisco’s F-Market street car, which recently had a collision on Market Street. Two users reported the collision — with 8 confirmations from fellow riders — 23 minutes before SFMTA’s official Twitter feed. In 2016, we will bring this technology to many more cities.

  Image of the F-Market collision being reported within Swiftly versus the official transit agency Twitter feed. As you can see, Swiftly users crowdsourced this incident 23 minutes before the agency broadcast it on Twitter.

Image of the F-Market collision being reported within Swiftly versus the official transit agency Twitter feed. As you can see, Swiftly users crowdsourced this incident 23 minutes before the agency broadcast it on Twitter.

SEAMLESS TRANSPORTATION

When transit issues inevitably occur, easy access to viable alternatives is essential to a multi-modal lifestyle. That’s why we integrated directions for walking, biking, and Uber (including real-time surge pricing). You can compare the travel time and price of your various travel options — based entirely on real-time data — without having to switch between multiple apps. We are continuously working to bring more choices to our users.

 Multi-modal trip planning that helps riders find the fastest and most affordable ways to get around town.

Multi-modal trip planning that helps riders find the fastest and most affordable ways to get around town.

COLLABORATING WITH CITIES

Our cities spend billions of dollars per year on transportation infrastructure with surprisingly little data to inform those investment decisions. Cities deserve more accurate and more complete information. We are making it a priority to work directly with cities and provide them with better tools.

We are pleased to announce that we’ve created a data-driven software suite that helps transit agencies improve their operational efficiency, make smarter infrastructure investments, and better engage their riders. We have already started piloting this system with select partners and plan to rapidly expand our footprint in 2016.

 Swiftly data platform for transit agencies and partners.

Swiftly data platform for transit agencies and partners.

JOIN US

We’re on a mission to make getting around town fast, affordable, and environmentally friendly. Working together, we can improve the way our cities move. We encourage you to give Swiftly a try!

The Swiftly Team