Traffic lights to stick around, despite the advent of self-driving cars.
It is a bit of a miracle that each day we drive around and encounter traffic signals that by the mere act of shining a light of differing colors, including a red light, gets us to all come to a peaceful stop, wait for one another to proceed, and then continue along on our own merry way.
The magical powers of red, yellow, and green.
A red light though doesn't have some kind of laser beam that will zap any cars that decide to not come to a proper stop.
There isn’t a hidden steel mesh that might suddenly pop-up from beneath the street to prevent cars from running a red light.
Instead, we stop at a red light because we are supposed to do so. Sure, you have the potential for getting a moving violation ticket if you don’t abide by the red light, which I’m sure reins in most of us, and there is also the danger of getting rammed by a car rightfully coming through the intersection, so we presumably do calculate our life-or-death chances and generally stop at red lights by weighing the risks of running a red.
All in all, it’s an amazing sight to see, namely the 225 million licensed drivers in the United States generally agreeing to accept and be managed by traffic signals.
I dare say it would be nearly impossible to get that many people to agree to any particular thing of any kind, yet by-and-large we all do agree when it comes to obeying traffic lights (most of the time).
Sadly, not everyone always does abide, and there is a significant portion of the annual 6.3M car crashes and 37,000 annual car-related deaths that can be attributed to violating a red light. Some violators do so out of a miscalculation, some do so because they didn’t see the red, some are drunk, some are car racing, some believe they are more important than the rest of us, and so on. Lots of reasons, but the end result can be disastrous.
Estimates are that there are 300,000 traffic signals across America.
I am pretty sure that a disproportionate share seems to be on the streets that I use to get to work.
There is nothing more frustrating than getting stuck at one red light after another, seemingly as though they are in cahoots with each other and wanting to keep me from getting to work on time. Admittedly, I try to get out of the red light abyss by purposely speeding up to try and catch the green light up ahead, but it often requires going so excessively above the speed limit that I cave in and drop my quest (though other audacious speeders near me seem to not care).
One interesting stat that often is quoted suggests that we spend perhaps 6 months of our lives sitting at red lights.
I’ve not added-up my red-light waiting times to see whether this stat is applicable. Given the at times unbearable on-hold at red lights I’ve endured, intuitively it sure seems like the grand total over a driving lifetime could be 6 months, or maybe it is more like 6 years where I live.
Traffic signals have evolved over time.
Plus, different countries have used differing colors or shapes for their traffic signals.
Since displaying only a color could be misinterpreted if you are color blind, there are traffic signals that also use shapes to distinguish the colors. Some traffic lights use a kind of emoticon etched onto the lenses as a means to distinguish the colors. In some places, blue is used as a color, while others use orange as a color for their traffic lights. You can use sounds for each of the colors, and you can make the lights flash or not flash, be bright or be dull, be stacked vertically or be shown horizontally, etc.
Here’s a claim that’s been making the rounds: Once we have AI-based true self-driving driverless autonomous cars there won’t be a need for traffic signals, and those then-useless traffic lights will all be quietly removed and turned into scrap.
Pretty much the notion is a bust, as will be next explained.
Why Traffic Signals Could Become Extinct
You might be wondering why pundits are asserting that the emergence of driverless cars will lead to the extinction of traffic signals.
The logic is that driverless cars are going to have V2V (vehicle-to-vehicle) electronic communications, allowing each autonomous car to electronically network with their brethren self-driving cars. When one self-driving car approaches an intersection, it will politely alert any cars coming toward the intersection to be aware of their upcoming passage through the intersection.
In theory, each driverless car will time its own approach to the intersection to allow other driverless cars to pass safely through the intersection.
Imagine that you and I were driving our own cars and had agreed to speak on our respective cell phones as we approached an intersection from different sides. I might tell you that in three seconds I’ll be into the intersection, and so you realize that you need to slow down to allow my passage, and then after I’ve cleared the intersection you proceed through it. The driverless cars are able to do the same, via electronic communication, in split seconds worth of timing.
Multiply this kind of civility by the dozens of driverless cars that approach the intersection, or ultimately the thousands upon thousands that would pass through any busy city intersection on any given hour or any given day. All of them would be programmed to be courteous to each other, doing so to avoid hitting each other.
Simulations trying to show the efficacy of this are pretty neat to watch. Simulated cars streaming toward intersections, and potential certain death, which nonetheless suddenly shift and tweak their driving to ensure none get hit and yet they all eventually get through the intersection. I doubt that a horde of ants could do as well.
In this federated autonomous dance, a traffic signal has no place, it is superfluous. Indeed, a conventional traffic signal probably could not do what the simulation is doing, since the faked cars are not waiting necessarily altogether and instead are each passing forward as they can, somewhat like the maneuvering of cars at a roundabout or traffic circle.
Might as well start hauling all those 300,000 traffic signals to the scrapyard.
Traffic Signal Not Quite So Dead After All
As Mark Twain might say, please don’t be so fast to write an obituary for traffic signals.
First, there are going to be for many years an ongoing mixture of human driven cars and driverless cars.
There are about 250 million conventional cars in the United States alone and those are not going to suddenly disappear when autonomous cars arise. We don’t even know if human driving will ever be cut-off or denied, though there are driverless car advocates that argue vehemently to get human driving expunged right away.
If there are human drivers, it likely makes sense to keep the traffic signals in place.
That being said, we might transform the nature of the traffic signal.
Rather than the signal being a set of lights, perhaps the traffic signal is converted over into an electronic beacon that sends out a non-visible indication of the red, yellow, and green status. Human drivers will then be required to have in their cars a piece of relatively simple automation that would detect the electronic signal and let the human driver know the status of the traffic indicator.
As such, the traffic signal wouldn’t even need to be placed into the middle of an intersection anymore. It could be on a nearby wall or building, or possibly be an edge computing “black box” emitting device bolted to a pole or fire hydrant. The signaling device could be much smaller than a normal traffic signal and be less likely to wear out (light-bulbs in traffic signals constitute a fair amount of maintenance effort).
Overall, it is anticipated that our roadway infrastructure is going to gradually be equipped with devices that can communicate via V2I (vehicle-to-infrastructure) electronic communications. For driverless cars, they will get V2I electronic messages from bridges that the pathway is closed, or get notified by streets that are being repaired, or receive messages from tollways that specify the cost to cross, allowing all autonomous cars to interact with the roadway infrastructure as needed.
Though this handy V2I arrangement will be feasible (once it’s put in place, at a substantial cost, mind you), let’s not forget about the other people that rely upon traffic signals. Pedestrians look at traffic signals. Bicyclists look at traffic signals. Motorcyclists look at traffic signals. And so on.
The use of lights is a pretty convenient form of communication. If the only emitting signal is electronic, you’d need to have a device on you or with you to detect it. Lights that are red, green, or yellow are readily seen and readily understood, not needing any added hardware or encumbrance.
Rather than junking traffic signals, it would seem more likely that we would gradually augment existing traffic signals with electronic beacons, which is already being done for various smart city efforts, and have a dual mode of communication, both the use of lights and the use of V2V and V2I.
Conclusion
There is the old joke about the person visiting a town and being driven in a car by a local.
The local comes up to a red light and darts through it.
The visitor is shaken and puzzled.
Upon arriving at another red light, once again the local zooms through the red.
Then, oddly, the local hits the brakes upon coming up to a green light.
Clearing his throat, the visitor asks why in the heck did the driver opt to run the red lights and now decide to halt at a green light?
The driver turns to the visitor and calmly explains, well, because his brother might be coming the other way.
Ha!
A great party joke.
Back to being serious.
Traffic signals are obeyed by a voluntary form of human conditioning.
Whether self-driving cars will do better than humans, assuming that someday there are only autonomous cars on the roadways, remains to be seen (presumably so).
Though self-driving cars won’t get drunk and won’t presumably try to road race, there is still the chance of system errors, glitches, and even cyberhacks that could undermine any kind of anticipated “perfection” in such matters.
On the other hand, I suppose that by the time we have achieved an all autonomous cars era, we’ll probably have flying cars, in which case, traffic signals might become extinct entirely since we won’t be on the ground anyway.
