Can we move toward electronic roads on the cheap?
What if traffic lights used wi-fi or radio to broadcast their decisions? Could cars use that data to arrive at lights just in time to catch the green? Could they talk back and actually vote on when the light changes?
Could cars use GPS, suspension sensors and wi-fi to report where the potholes are?
There's work being done on enabling cars to travel closer together and reduce the amount of drag that most cars have to overcome, but isn't there lower risk, low hanging fruit to pick?
Can you think of any creative ways to deploy the great technologies we have now to make concrete more efficient?
(
6 comments
| 7 points
| on HN
)
Many city planners believe that stopping traffic frequently makes the city safer. This isn't true, of course. But, it seems true and it's easy to believe. If a city timed major traffic ways properly, it's been shown that accidents are reduced. But, facts are so inconvenient. Rather, lights are timed poorly which causes greater pollution and tension. The drivers, seeking a more optimal path, start taking side roads and neighbors complain. So, the city, rather than improving the efficiency of their roads, make the neighborhood streets less attractive by adding stop signs and speed bumps.
The other factor is the profit motive. Frustrated drivers who have limited options for side streets become increasingly likely to slip through changing lights or speed between signals. The result is, of course, greater revenue for the city. You'll note, this is also a completely unsafe situation. Ideally, our municipalities are there to make things better, easier, safer. Here, they're not -- yet, no one seems to complain.
What you can do, rather than introducing yet another optimization agent (humans do this pretty well), is organize a social movement to make your streets safer through applying demonstratively better solutions with existing technology. For instance, why do demand lights initiate an immediate change rather than simply queuing the change durning the next light cycle? Negligence and miserliness.
Moreover, in places where the lights are influenced by the cars around, I think it's usually pretty easy to tell if a light is going to change soon.
But, I do think we will start seeing a lot more tech in the automotive future. I think it's the newest prius that has an auto-sensing cruise control, to adjust the speed to the car in front of you. I've been wanting that for years. Also, I think it's the first step in integrating autonomous drivers on the road.
I've also heard about work on a "train" like vehicle that other cars could attach to and be dragged along. Idea is that it'd be more efficient, plus the drivers would get to relax on their commute.
Like you mentioned about communicating where potholes are, I think all cars should be alerted to emergency personnel on the side of the road well before they reach them. I think maybe some radar detectors do something like this, maybe as a side effect. Something built into GPS would be awesome though.
But what if one had sturdy beams instead of wires that could guide the trolley-poles, and indirectly the driver to stay in-lane. More like a third-rail, that is 4-5m above the road.
These trucks and buses could then have a weaker combustion engine to use off electricity, but perform the majority of the trip, on the freeway, on electricity.
I understand that not all countries have a near mix of 50% hydro and 50% nuclear electric power to make this decrease pollution.
I can imagine hooking up a traffic light with both a wi-fi device and cameras to measure traffic flow. Traffic light A could communicate with traffic light B to let each one know how much traffic was on the way to the other light, so that it could adjust its programming. If you did this on a large scale, I would expect it to reduce traffic by a lot.
Of course, I don't know if the reduction in traffic would be greater than the cost of equipping all traffic lights with wi-fi devices and cameras.
But if you can measure traffic flow passing through one intersection, can't you measure traffic approaching another?
It seems that the way to start implementing would be to improve nodes first. Connecting them seems a bigger, more political job.