It’s Possible That Google Made a Mistake

with its AI Traffic Light Project

 

 

 

 

“Traffic has so many uncertainties.”

Although it’s unclear exactly how beneficial Google‘s covertly implemented AI traffic signal initiative is, the company says it streamlines processes and cuts down on wait times at stoplights.

As explained by Scientific American, Google‘s Project Greenlight was first tested in 2021 and is based on an algorithmic approach that aims to replace the other two main traffic signal control systems now in use.

Urban planners have long struggled with traffic planning and control, as the systems rely on either manually adjusted and fixed light changes or below-ground sensors that tell the devices how many automobiles are at a specific intersection.

Here comes Project Green Light, which trains the system to predict when traffic will get worse and adjusts stoplights accordingly. It does this by using a model of traffic known as adaptive or responsive traffic that is fed data from Google Maps.

Recently implemented in Boston, Project Green Light has been utilised with differing degrees of success in Seattle, Manchester, England, and other places.

Although it’s not the worst thing Google has ever connected, anyone who has ever been in a collision understands that unexpected things can occasionally happen on the road and cannot always be foreseen by an algorithm.

While officials from other cities who employ the Google traffic light AI noted in interviews with the magazine that its recommendations frequently fell short of the target, Boston brags about its relationship with the project.

For example, representatives from Manchester, an English city, told SciAm that their traffic engineers routinely disregarded Project Green Light’s recommendations since they weren’t very good.

The algorithms neglected to consider things like bus routes, so the same programmers had to manually alter stoplights to prioritise them or divert commuters from driving through residential neighbourhoods.

Even Mariam Ali, a spokesperson for the Seattle Department of Transportation who oversaw Project Green Light and who was generally complimentary of the program, admitted to its shortcomings.

Ali stated to the magazine that the city had “seen positive results,” but she also acknowledged that a traffic shift that Google had proposed had to be reversed since it “did not result in a net benefit.”

Traffic control expert Aleksandar Stevanovic of the University of Pittsburgh told the magazine that while it’s “great that Google is working” on putting high-tech solutions into place to solve the bewildering problems that arise from traffic signals, human decision-making will always be crucial.

“There are so many unknowns in traffic,” Stevanovic said. That’s “not rocket science,” he continued, adding that it’s actually “more difficult.”