Be Resolute To Reduce That Nighttime Artificial Light Pollution, Which AI Self-Driving Cars Can Help Solve (But Doing So Will Be Tricky)

Beijing traffic congestion



Let there be darkness.

That is the potential catchphrase for those that are concerned about nighttime light pollution.

More formerly known as Artificial Light At Night (ALAN), there is an ongoing bruhaha that our modern way of living is generating way too much light during the evening darkness. It is an ongoing issue and the amount of such pollution is likely to keep on increasing due to further industrialization and expansion of societies into additional geographical areas.

In short, you can expect more light to be emitted in existing populated areas, along with nighttime light being unleashed in regions that had so far not been especially well lit due to insufficient means or lack of a light-producing populace. When you start adding more office buildings, more homes, more cars, more lampposts, and the like, this all translates into a tsunami of unbridled light at night.

You might be puzzled as to why the mere shepherding of artificial light is considered a pollution monstrosity.

One obvious facet is that you cannot see the stars at nighttime, or they are otherwise generally blotted out of view by the abundance of overwhelming artificial light.

How does this happen?

Well, there is the artificial light that comes from say a skyscraper, considered to be a point source of such light, and there is the skyglow type of light, proffering a combined effect that muddles together the thousands upon thousands of point source lights. I’m sure you have noticed this effect whenever taking a flight over a sizable part of the country and observed the glowing lights from any big city. Particularly striking is the evening glow around Las Vegas which is a look-here lights-on-steroids city plopped down into the middle of an otherwise utterly darkened desert.

Some researchers estimate that about 80% of the global populace lives in light-polluted areas. This is somewhat concentrated geographically because there are vast parts of the world that are relatively uninhabited or sparsely so. Research suggests that perhaps one-fourth of the world has those polluted nighttime skies.

Besides undercutting the ability to readily see the stars, there are other quite notable qualms about ALAN.

For example, some assert that artificial lighting is impacting the everyday life cycle of nature. Animals that are cued naturally to act at nighttime are at times taken aback by the appearance of artificial light, and likewise, animals that focus on daytime lighting are seemingly triggered to act in the evening when they normally would not do so. It is believed that humans can also be impacted by all of the artificial light at night, perhaps being linked in some ways to poorer sleep quality amid disturbances to circadian rhythms, might spur or promote impaired memory capacity, can possibly foster obesity, and so on.

Various recommendations have been made about corralling this out-of-control artificial light abundance. Experts ask that artificial lighting only be used when fully justified. Artificial lighting should not occur simply by default and indeed there ought to be adequate attention allotted toward the polluting properties that can arise. When feasible, artificial lighting should be kept to the lowest usable intensity, it should be shielded and guided toward whatever the intended lighting purpose requires, etc.

You might not realize that you are probably a nighttime artificial light polluter.

How so?

When you go for a drive at nighttime, your headlights become a source of artificial light pollution. That being said, do not foolishly think that you should not be using your headlights. The use of your headlights is a vital safety measure and wholeheartedly needs to be used. You have to see the roadway to properly drive a car. Without headlights, you’d be driving in the dark, dangerously so, and undoubtedly the outcome would be disastrous.

You might have noticed from time-to-time cars driving along in traffic that do not have their headlights on, even though it is completely nighttime and all the other nearby vehicles have their headlights on. We usually assume that the driver might be intoxicated and not aware that they are driving without their headlights employed. Oftentimes, police will pull over a vehicle that doesn’t have its headlights on and do so to find out why the driver is driving without this needed safety capability engaged.

During times of dusk and sunrise, you’ll often see that some cars have their headlights on and some do not. The somewhat gray aspect of whether to turn on your headlights when there is a modicum of natural light still available or emerging is one that drivers make a judgment call about. Of course, most modern cars have an automated feature that senses the available lighting (or lack thereof) and will turn on the headlights for the driver, thus the driver doesn’t have to take into account per se the headlight engaging chore.

Why all this talk about artificial light pollution?

A reader of my column recently sent me an intriguing question about the future of cars and the matter of artificial light pollution. As you likely know, cars will gradually become self-driving cars. Those are cars in which an AI-based driving system acts as the driver and there isn’t a human driver at the wheel (see extensive coverage in my columns at this link here).

The question was this: Will the advent of AI-based true self-driving cars potentially allow for the turning off of headlights at nighttime due to the self-driving cars orchestrating automatically their movement and traffic flow?

This posits the conjecture that perhaps self-driving cars will electronically communicate their positions and align to avoid each other and proceed seamlessly as though like the flowing of water in a stream. The hook regarding headlights is that presumably there won’t be a need to use headlights when those self-driving cars are all dancing together with aplomb. Just imagine, if it was possible to turn off the headlights and yet still had utter safety of traffic (in this imagined self-driving vehicular vision), there would be a substantive reduction in the artificial light pollution (at least due to headlights).

Let’s consider whether that makes sense.

Before we jump into the fray, it will be helpful to clarify what is meant by the notion of AI-based true self-driving cars.

Understanding The Levels Of Self-Driving Cars

As a clarification, true self-driving cars are ones that the AI drives the car entirely on its own and there isn’t any human assistance during the driving task.

These driverless vehicles are considered Level 4 and Level 5 (see my explanation at this link here), while a car that requires a human driver to co-share the driving effort is usually considered at Level 2 or Level 3. The cars that co-share the driving task are described as being semi-autonomous, and typically contain a variety of automated add-on’s that are referred to as ADAS (Advanced Driver-Assistance Systems).

There is not yet a true self-driving car at Level 5, which we don’t yet even know if this will be possible to achieve, and nor how long it will take to get there.

Meanwhile, the Level 4 efforts are gradually trying to get some traction by undergoing very narrow and selective public roadway trials, though there is controversy over whether this testing should be allowed per se (we are all life-or-death guinea pigs in an experiment taking place on our highways and byways, some contend, see my coverage at this link here).

Since semi-autonomous cars require a human driver, the adoption of those types of cars won’t be markedly different than driving conventional vehicles, so there’s not much new per se to cover about them on this topic (though, as you’ll see in a moment, the points next made are generally applicable).

For semi-autonomous cars, it is important that the public needs to be forewarned about a disturbing aspect that’s been arising lately, namely that despite those human drivers that keep posting videos of themselves falling asleep at the wheel of a Level 2 or Level 3 car, we all need to avoid being misled into believing that the driver can take away their attention from the driving task while driving a semi-autonomous car.

You are the responsible party for the driving actions of the vehicle, regardless of how much automation might be tossed into a Level 2 or Level 3.

Self-Driving Cars And Artificial Light

For Level 4 and Level 5 true self-driving vehicles, there won’t be a human driver involved in the driving task.

All occupants will be passengers.

The AI is doing the driving.

One aspect to immediately discuss entails the fact that the AI involved in today’s AI driving systems is not sentient. In other words, the AI is altogether a collective of computer-based programming and algorithms, and most assuredly not able to reason in the same manner that humans can.

Why this added emphasis about the AI not being sentient?

Because I want to underscore that when discussing the role of the AI driving system, I am not ascribing human qualities to the AI. Please be aware that there is an ongoing and dangerous tendency these days to anthropomorphize AI. In essence, people are assigning human-like sentience to today’s AI, despite the undeniable and inarguable fact that no such AI exists as yet (see my discussion at this link here).

With that clarification, you can envision that the AI driving system won’t natively somehow “know” about the facets of driving. Driving and all that it entails will need to be programmed as part of the hardware and software of the self-driving car.

Let’s dive into the myriad of aspects that come to play on this topic.

The first aspect to consider is whether self-driving cars will possibly be orchestrated by a centralized means that will coordinate their actions. This is a commonly expressed notion by some pundits. A centralized controller would attempt to maximize the flow of traffic and seek to reduce or entirely eliminate any semblance of congestion or traffic snarls.

Generally, though that might work in a simulation, you would be hard-pressed to make this contrivance viable in the real world.

Part of the reason that some believe this could work is that the AI driving systems will presumably take any strict orders about where to drive, what speed to drive at, and so on. Unlike human drivers that would likely balk at a centralized command that tells them what to do, the AI driving systems would simply carry out whatever instructions were provided (well, that’s assuming that the AI was programmed to do so).

Unfortunately, there are a lot of gotchas and traps involved in trying to bring to fruition that kind of centralized control. In brief, this would require an ostensibly flawless electronic communications network that would transmit the car-going instructions and ensure that each self-driving car got a timely indication of what needed to be done. Even a split-second delay or if there was noise on the transmission could lead to a calamity in this rather farfetched scenario.

Another false assumption underlying the Utopian-like setting is that the roadway could have all sorts of imperfections and considerations that this centralized command might be unaware of. Suppose a piece of debris suddenly slips onto the roadway. Meanwhile, the centralized command is telling the self-driving cars to proceed ahead. Wham, a self-driving car hits the debris and then a cascading series of car crashes occur.

Yet another issue that sometimes is forgotten entails the aspect that we aren’t likely to have only self-driving cars on our roadways, namely there will also be human-driven vehicles too. The emergence of self-driving cars will be gradual and take many years to become prevalent. During that time, likely many decades, there will be a mixture of self-driving cars and human-driven cars in traffic. In fact, some human drivers insist that they will never give up driving, until the day that you pry their cold dead hands from the steering wheel.

The point is that you might be able to get the self-driving cars to abide by the centralized commands approach, but one wonders whether human drivers will be willing to do so. Also, if a human driver decided to divert from the centralized command, doing this while in the middle of a sea of self-driving cars, it could once again cause quite a shebang akin to the example earlier about the appearance of debris. The human-driven car would be where it is not expected to be, and the nearby self-driving cars might ram into the human-driven car or take evasive action, upon which the centralized command might be delayed in being informed.

Let’s then reasonably agree to set aside, for now, the centralized command notion as a means of trying to turn off those self-driving car headlights.

We might also amicably agree that headlights are a necessary enabler for vision-based driving.

Humans primarily use their vision as the key or core sensory apparatus needed to drive a car. When there is sufficient natural lighting, our eyeballs or ordinary vision can see the road. When there is a lack of sufficient natural lighting, our eyeballs cannot fully see the road. We, therefore, use artificial lighting to support our vision-based driving.

Lampposts near a roadway are handy to artificially light the roadway and aid our eyeballs as we drive. In theory, we could dispense with headlights if there was sufficient lamppost lighting on every nook and cranny of wherever you might want to drive. The beauty of headlights is that you bring them with you wherever your car goes. You don’t need to depend upon any external source of lighting such as lampposts, and instead, you are carrying with you the ability to project artificial lighting when needed, where needed, as needed.

This brings up some interesting considerations worth pondering.

One aspect entails whether self-driving cars need to use vision or perhaps could be devised to use other sensory types.

As you might know, self-driving cars are usually loaded with video cameras and those cameras are collecting data about the driving scene. The AI driving system interprets the images coming from the cameras. Using Machine Learning (ML) and Deep Learning (DL), which are computational pattern matching techniques, the AI driving system tries to ascertain where other cars are, where pedestrians are, and other objects within the range of the self-driving vehicle.

In addition to video cameras, many of the self-driving cars are equipped with radar, LIDAR, thermal imaging, ultrasonic units, and other kinds of sensors. If we wanted to do away with the headlights, we might do so if we could rely entirely on the other types of sensory devices during the driving task. Thus, by somehow excising the vision elements of driving, a self-driving car would no longer need to light the way via headlights.

Sorry to say, this seems untenable in today’s world. Not using vision is just not in the cards for now. The opposite is actually being generally pursued, wherein some automakers and self-driving tech firms are pushing further ahead on vision, aiming to make vision even more crucial, and downplay the role of the other sensory types.

Scratch for now the idea that you could eliminate the need for vision in self-driving cars as a scheme to get rid of the headlights and reduce the artificial light pollution at nighttime.

Let’s try a different angle.

Have you ever noticed that when you are driving adjacent to other cars that have their headlights on that you can oftentimes see the road due to their headlights as an afterglow or spillover effect? I’m sure you’ve noticed this. I doubt that you considered turning off your headlights since you did not know how long those other nearby cars would be near you. No sense in taking a chance of having no lighting when those other vehicles change lanes, slow down, or otherwise change their traffic positions.

But with self-driving cars, there is a chance of knowing where and when the other nearby self-driving cars will change their positions. Using V2V which is vehicle-to-vehicle electronic communications, self-driving cars will be sending messages to each other. This might include for example the possibility of one self-driving car detecting debris on the roadway and then alerting electronically other self-driving cars to watch out for the obstruction.

Okay, we might then do this.

Whenever self-driving cars are near to each other, while at nighttime, they would electronically communicate about the use of their headlights. The headlights from one self-driving car could be illuminating the roadway sufficiently for another self-driving car nearby, such that the self-driving car getting the spillover could turn off its headlights momentarily. This might last for a few seconds or maybe minutes, depending upon where the self-driving cars are going and the nature of the roadway being used.

Upon an electronic communication from the self-driving car with its headlights on that, it is intending to shortly move away from the other headlight “borrowing” self-driving car, the self-driving car without its headlights on would need to turn on its headlights. All of this could happen very quickly. Electronic messages would be zipping around and you would see a seemingly oddball-looking scene as the self-driving cars all were turning off and on their headlights. Envision a potential twinkling effect similar to those sparkling lights on a Christmas tree.

In theory, maybe this would balance out that perhaps on the average you would have one out of every two self-driving cars using its headlights entirely on a driving journey, or maybe just one out of every ten, or every hundred, etc. It would depend greatly on where these self-driving cars are, how close they get to each other, and a myriad of other factors.

Would the logistics of this be worthwhile in terms of the reduction of ALAN?

A back of the napkin analysis suggests this would be quite a reach of achieving any significant dent in the artificial light pollution matter. We also need to once again consider how this might impact nearby human drivers. And so on.

Do we give up on this?

No!

Consider another possibility that involves throttling the amount of light being emitted from your headlights.

Imagine that headlights were able to showcase a beam of light that could be ratcheted between a numeric range of perhaps one to ten (a zero would mean the headlights are not on). The headlights at a setting of one were barely on. Headlights at a level of ten are super high beams.

For a human driver to try and continually throttle or adjust their headlights would seem onerous and a severe distractor from the driving task. On the other hand, this would be a relatively easy task for the AI driving system to computationally perform and assuredly not distract from the driving activity. The video cameras or other akin sensors would detect the needed level of lighting and the headlights would be electronically commanded to adjust accordingly. This could happen in split seconds and be a routine function of the vehicle.

Here’s how this might help to reduce artificial light pollution.

Suppose that most of the cars on the roadways are eventually self-driving cars. Those self-driving cars can quickly and efficiently adjust the amount of light needed from their headlights. If they are all doing this, it could add up to a lot less light being emitted, in total and on average. Assuming that safety is not compromised, this could be a means of reducing the ALAN aspects as spurred by car headlights.

Would the reduction in ALAN be worth the tradeoffs of employing such a feature?

That’s an open question.

Conclusion

There are additional twists and turns on this topic.

Some assert that we might use cars at nighttime more so than we do today as a result of the advent of self-driving cars.

This seems like a sensible piece of guesswork. Self-driving cars will be running 24x7 and there won’t be a need for a driver to take breaks, rest, or grow weary of driving. People that want to go from point A to point B might be more willing to do so at nighttime than they are today, especially since the self-driving cars might be less costly for riders than when using human drivers (we don’t know this to be the case, but it is highly speculated to occur).

The bad news then is that we might produce more artificial light pollution due to the use of self-driving cars. Not because the self-driving cars are producing more light, but because they will be used at nighttime in larger volumes than potentially human-driven cars of today are used. Yikes, it seems like self-driving cars are going to contribute to ALAN and not reduce ALAN if you buy into the foregoing logic.

Wait for a second, there could be a countervailing aspect.

Some claim that we will have fewer cars on the roadways as a result of self-driving cars becoming available. People will no longer own cars and will instead use ridesharing self-driving cars. You could therefore contend that though there might be more nighttime trips, perhaps it might occur via a lesser number of vehicles, and ergo the amount of light pollution is the same or reduced in comparison to human-driven and personally owned vehicles.

Whew, that seems a potential relief.

The thing is, we are all pretty much in the darkness right now about how self-driving cars are going to turn out. There is no magic light that shows what way self-driving cars are going to go (a bit of an extended pun, thanks for noticing).

Aristotle famously said that the light of day is followed by night, as a shadow follows a body. We are unlikely to be able to do anything directly about nighttime, but we can do something about the artificial light that we use when it is dark out. The headlights topic is pretty low on the priority list right now for attaining self-driving cars. Upon eventually giving due consideration to the headlights topic, it would tend to imply that we’ve already solved the zillions of other unsolved more pressing problems facing the advent of self-driving cars.

No comments

intech company. Powered by Blogger.