IBM Will Go All Watson On Your Commute, Keep You Out Of Traffic

Imagine a world where no one ever gets stuck in traffic--where cars have built-in sensors that can predict where and when future accidents will occur, keeping commuters out of harm's way. That's never going to happen.

But IBM, Caltrans, and the University of California at Berkeley are working on the next best thing: personalized commuter forecasts that analyze the traffic on individual routes, warn drivers of the rush-hour madness to come before they leave the house, and suggest new travel plans (including alternative forms of transportation).

The technology is still in the prototyping stages (a group of IBM employees are testing it out), but one day it might go something like this: Your phone's GPS keeps track of your daily route to work, courtesy of an IBM smartphone app. Before you head out in the morning, you get a notification on your phone telling you that every part of your commute is stacked with cars. Instead of slogging through the traffic, your phone recommends that you drive halfway to work, park in the BART parking lot, and take the subway system the rest of the way. If you leave now, you'll make your way through traffic just in time to catch the next train to work.

IBM hasn't invented any crazy new technology to make this work. The system uses the thousands of road sensors installed by Berkeley and Caltrans over the past decade to figure out how many people are driving and how much traffic is in a given lane. These same sensors are present in major metropolitan areas across the country.

"Analytics enable us to take all of those data points and build a historical database for how traffic behaves and make real-time predictions. So if one particular slowdown in a particular spot correlates with slowdowns in other spots, we can estimate that corresponding slowdowns will occur right after that," explains John Day, Program Manager for IBM Smarter Traveler. The public transportation part of the equation also uses already-existing sensors to predict when trains will arrive.

Beyond helping people make it to work without wanting to cry from frustration (and saving gasoline), IBM has another motive for creating personalized commute forecasts: cash. The company speculates that one day, it could sell the system to transportation partners and run the backend. Those partners could even implement a system paid for by advertisements. We assume that this might allow merchants along a travel path to advertise special offers for commuters. And there you have it: A blessedly easy commute in exchange for IBM knowing your exact location.

One caveat: we know that lots of you don't have alternative routes that you can take to work, or the flexibility to hop out of your car onto public transit. If that's the case, we're sorry. Not even the eggheads at IBM can save you from your crappy commute.

Reach Ariel Schwartz via Twitter or email.

Add New Comment

0 Comments