February 20, 2013 By Naveen Lamba, associate partner, Intelligent Transportation at IBM Global Business Services
Anyone who has suffered through stop-and-go traffic has at least once daydreamed of soaring in a Jetson-style flying car to their final destination, avoiding altogether the endless line of taillights.
Though flying cars are far from reality, self-driving cars have emerged as a potential high-tech cure for congestion. Researchers are racing toward a goal where smart cars take over the burden of driving. Tantalizing as the idea may be, it will be a decade or more before such technologies hit the mainstream.
But don’t write off high-tech fixes for congestion just yet. A growing constellation of technologies — built into our cars, wired into our smart phones, and being built into our streets and highways — will push back the rising tide of traffic.
Congestion costs to the U.S. economy are $121 billion — comprised of gallons of gas burned while idling, hours lost in traffic, and goods delayed in delivery, according to the Texas Transportation Institute’s annual survey of traffic’s economic toll.
The study finds that in many regions, rush hour snarls can extend a journey up to four times its traffic-free duration, underscoring that traffic woes are widespread and stubborn. Most of this year’s top 10 most traffic-snarled cities are repeat offenders, topped by Washington, D.C.
Cash-strapped cities and states are falling behind in maintaining existing infrastructure, let alone constructing new roadways; one single mile of highway construction can cost tens of millions of dollars.
It’s premature to concede the fight to congestion. While technology may never wholly prevent the toll of traffic jams, solutions are evolving to improve the daily commute. Unlike driverless cars, these technologies are already coming on line. A modest investment in technology can make a quick impact on reducing congestion.
It starts with data. Our roadways, in addition to carrying millions of vehicles, yield rich veins of information that can be used to understand, analyze and reshape traffic.
For instance, the density and movement of cell phone signals — scrubbed of personal information — is an ideal indicator of traffic density on given roadways.
Likewise, more cars and smartphones are equipped with two-way GPS devices, able to retrieve data on road conditions and interact with remote servers to plan alternate routes.
Public and corporate entities play a role too. In some regions, taxies, delivery trucks, city vehicles and other large fleets share transport data. Officials can also monitor traffic flows using digital sensors in roadways, overhead video cameras, and automatic toll collection systems such as EZ-Pass.
Translating this big data into helpful guidance requires large-scale computational power, running complex analytics systems capable of scouring the data in real time. The goal: to instantly assess how traffic is flowing and predict congestion before it happens, using historical records of past traffic patterns.
This process is taking shape in a handful of sites. In France, Germany and The Netherlands, for instance, transportation planners have used IBM cloud-based analytic software and systems to predict how fast traffic will be flowing up to 45 minutes later.
By seeing into the future, traffic managers can respond before trouble starts — altering traffic-light timing, opening up more toll lanes, or pre-positioning police and tow trucks — to help minimize the congestion. The software can also help determine which of these or other steps will deliver the best benefits.
Smart traffic algorithms are beginning to tap a potentially infinite source of congestion indictors: driver opinion. Software agents can scour social networks such as Twitter and Facebook for public posts that include “traffic,” “congestion” and “delay”. Sudden spikes in these keywords in a given geography are a bright beacon to pinpoint delays.
These early efforts at harvesting data to alleviate traffic are just the beginning. The more data made available to these systems, the greater our potential to more accurately predict and prevent congestion.
More data by itself won’t help — commitment to make better use of this information is required. Resources to better understand these data flows must be recognized with equal or more importance as the traditional infrastructure investment. Few other remedies offer as much promise to cost-effectively get more from today’s infrastructure.
Visionaries will hopefully never give up on their chase of Jetson-ian technologies. Until then, the more we understand congestion, the better we can reduce traffic, shorten commutes and improve the livability and economic viability of our cities.
You may use or reference this story with attribution and a link to