By arvind |Email | Jul 31, 2018 | 9303 Views

While the intersections sit empty you sit still in your car waiting on that red to go green, in that very moment that is the only go green motto you can think of. AI can make that frustrating inefficiency of traffic a thing of past. A meme to remember only the 2000s kid's will remember". Just think there could be a tie in the near future when there is no traffic. Traffic could be something that is taught in the books of history's kids there was a time when cars used to just sit in lanes for no apparent reason". You could brag to the next generation about traffic life is easy for you we used to get stuck in a thing called traffic when we used to travel". Doesn't just reading this makes you happy, get me those days tomorrow.

We humans are here to stay until an Ultron losses his mind and goes on a mission of killing humans. Humans are not going anywhere and neither are the traffic lights. But smart traffic lights and smart cars can make getting around smoother for both drivers and pedestrians, researchers are taking steps towards a future like this. Companies are working on why traffic congestion is happening collecting data from hundreds of sensors and learn how to manage it. Rapid Flow Technologies which began as a Carnegie Mellon University project is testing surface traffic management system in its Pittsburgh neighbourhood. An AI based system called Surtac has reduced waiting times at traffic lights by as much as 42% in the Pittsburgh neighbourhood.

The traditional traffic signals commonly change on a fixed schedule. Some are coordinated with those at the next intersection. It is so fixed that the internal clocks of humans has started picking on it, people start honking 5 seconds in advance before the green light, remember go green. How advance a traffic light is, well some can sense a car waiting and adjust the timing, as futuristic as it gets.

For years apps like Google maps have been capturing traffic information by crowdsourcing their data, monitoring it, getting accident reports traffic reports and relaying that information to other users from other users. Despite all the possibilities offered by AI for network traffic control, there are a number of new challenge including adaptive scheduling of artificial computing, cooperation of heterogeneous intelligent schemes, and computational complexity. 

AI will not only help reduce the traffic make traffic lights smarter but will also help reduce auto emission because the less the time a car spends on the road.

Source: HOB