Dump Truck Accident Pours Liquid Asphalt From Overpass Onto Vehicles Below, Providing Lessons For AI Self-Driving Cars
Asphalt is all around us, that’s for sure.
Our highways are typically made of asphalt. Streets are made from asphalt. Parking lots. Airport runways. Even tennis courts and the very rooftops on our homes are oftentimes reinforced or entirely composed of some quite handy dandy asphalt.
Asphalt that is used to construct these various everyday elements is usually trucked to the location that needs the asphalt for building purposes.
During transit, the asphalt is normally kept at around 300 degrees Fahrenheit. This ensures that the potent mixture remains fluid-like and can be readily poured into place. To further enhance the liquidity of the material there is an added ingredient of either kerosene or diesel oil. And, you might be surprised to know that some dump trucks are outfitted to purposely feed their heated engine exhaust into the body of the asphalt pile, providing an extra means of keeping the asphalt pliable while being transported.
You’ve undoubtedly seen dump trucks carrying asphalt and witnessed them from time to time as they crisscrossed your city streets on their way to a nearby construction site.
I’m betting that you’ve even been on a freeway and had the distinctly unfavorable honor of driving directly behind a truck loaded with some of that roasting asphalt. You can usually smell the stinky distinctive odor emanating from the loaded-up vehicle. Furthermore, when the truck is full and carrying a heavy load of the material, there is a high chance that the bulky vehicle will be moving along relatively slowly in comparison to the rest of the traffic.Be forewarned that the fumes of the steaming asphalt can be dangerous to your health. The odds though of getting harmed by a few whiffs of the stuff as it is being carted along seems ostensibly remote. Another potential danger consists of getting hot asphalt onto your skin. That can be horrendous and besides the burning aspects, there are also the chances of the material being absorbed into your skin.
A word to the wise, do not let small children go play with freshly deposited asphalt. Adults are at times tempted to let kids go ahead and mess around with a newly laid piece of street or parking lot asphalt. Do not do this.
Enough said.
Returning to the notion of at times finding yourself driving behind a truck that is loaded up with fresh asphalt, one consideration that might pop into your noggin is the oddball chance that the asphalt might somehow be bumped out of the truck and onto the roadway in front of you. Perhaps the truck strikes a big piece of debris causing the truck to take a bit of a jump, and the next thing that you know there is a slushing of this liquefied asphalt plopping onto the freeway or highway.
Yikes!
What would you do?
I suppose it depends upon how close you were to the inadvertently slushed material and how much of it was dumped out.
The hope would be that you could just continue straight ahead and straddle the material so that your car harmlessly passed over the yucky stuff. With luck, your tires wouldn’t get mired in the foul mixture.
For a brief instant, you would have likely been immensely startled and dreading the worst of outcomes. Upon zooming over the toxic mess, you would glance in your rearview mirror and be thankful that by your lucky stars the potential hazard was evaded.
Suppose though that the volume of the inadvertently dumped asphalt was much larger. In that case, you might try to veer into another lane and avoid the scary mound of noxious slush. The act of sharply switching lanes might have its own dangers. If there are cars in those lanes you then risk the chance of colliding with those vehicles.
Which would you prefer, potentially striking other nearby moving vehicles or plowing into a large pile of heated asphalt?
Doing so at freeway speeds such as 60 miles per hour.
None of us ever wants to face that kind of worst-of-two-evils choices.
It is the proverbial classic case of being between a rock and a hard place.
You might be right now thinking that the odds of being behind a dump truck that has a full load of heated asphalt and having any of the material falls onto the roadway in front of you is just about nil. In other words, sure, there is some wildly bizarre chance that this could someday happen, but it seems as unlikely as having space aliens beam you up into a visiting spacecraft from outer space.
Well, I guess you ought to start looking skyward.
Why so?
A dump truck carrying a fresh load of asphalt was on an overpass above a freeway and the driver accidentally collided with the side railings. In doing so, the dump truck pitched to its side and the asphalt summarily poured onto the traffic below.
Take a moment to let that scene soak into your mind (sorry, bad pun).
Maybe the phrasing of saying that it poured out is not an entirely apt way to describe the ghastly act. The asphalt gushed out. It flowed at what seems like the speed of light (not really, but you know what I mean). One second the asphalt was wholly contained in the dump truck, and merely a minuscule couple of seconds later it was all drenching down onto the freeway below.
In the earlier discussion, the scenario envisioned was that a dump truck in front of you might inadvertently unleash some asphalt. That would be bad, but at least you would undoubtedly see it happen. Your eyes are normally riveted on the traffic ahead of you.
How often do you look upward for potential dangers to your car?
Not very often.
We tend to assume that it is unlikely to have something plummet from above and strike our vehicles.
Indeed, I was driving past a golf course one day and got caught unawares by a loud smack of something that whacked into my front windshield, cracking it quite a bit. Yes, this was an errant golf ball. Some golfer had a really bad swing and managed to send a golf ball out of the golf course and into traffic. And into my car.
Just my luck.
I admit that I didn’t see it coming. Please be aware that the wayward golf ball had gotten over a tall netting that was surrounding the golf course. I mention this aspect because some friends were at first surprised that I was somehow so clueless to not see the golf ball before it struck my car. They made me out to be a lousy driver that I didn’t see this tiny speck that just so happened to archway above my normal gaze and plummet at bullet-like speeds down into my windshield.
Give me a break.
Anyway, imagine that you are driving along on a freeway, minding your own business. You are attentive to the traffic around you. You are watching the traffic ahead of you. Let’s assume you aren’t watching cat videos and nor trying to text on your smartphone while driving. You are exclusively in the mental zone of being a courteous and mindful driver.
A torrent of steaming asphalt suddenly appears out of nowhere, flowing in front of you as though a raging river of asphalt was unleashed. Keep in mind that the asphalt was coming down from the overhead overpass. It is akin to a waterfall, though consisting of dangerous, toxic, stinky, burning, slimy, mushy, and altogether a frightening tsunami of hot asphalt.
That’s when you know that fate is not being kind to you.
There is a viral video on social media that captures the whole scene (the event occurred in Shanghai, China). A stunning aspect is how quickly the event happened. Within two or three seconds of the dump truck hitting the railing, it gets pitched to its side and the asphalt rains out.
Wham. Slush. Bam.
There isn’t some lengthy time involved about the asphalt slowly oozing over the railings. Nope. In nearly the time that it takes to snap your fingers, the asphalt goes from the bed of the dump truck and immediately and without any hesitation down onto the traffic below. By a miracle of sorts, only four cars got caught in the mess and just two people got injured (according to news reports).
Things could have been a lot worse.
One thing to ponder is whether you might have been able to notice the dump truck going across the overpass, which certainly seemed visible from the freeway below. Maybe you would have realized what destiny was aiming to unleash.
We all drive on highways and freeways that go underneath overpasses. This is common.
Do you routinely look up at overpasses?
Maybe somewhat yes, maybe somewhat no.
The odds are that you visually see the overpass but do not give it any undue mental focus. Without any direct noodling, your mind is presumably subconsciously looking for anything only out of the ordinary.
For example, if a person is standing up there on the overpass and facing the freeway below, this might catch your eye. There have been reported cases of miscreants that opted to toss rocks or bricks onto passing cars (a decidedly criminal act). Knowing that this can happen, you would certainly be observant of someone suspiciously standing still on an overpass and facing the traffic.
Thankfully, those outlandish happenings are a rarity.
All told, probably 99% of the time, you do not give any explicit attention to overpasses. Unless the overpass is crumbling from old age, or maybe there is a parade of specially decorated cars going across it, the roadway above you is stridently out of mind and not of particular concern. There is enough other stuff to concentrate on, such as the vehicles in front of you and to your sides.
I would further submit that even if you perchance noticed a dump truck going across the overpass, that too would not cause you any mental stir. Trucks and big vehicles are constantly using overpasses. If you continuously took your attention away from the hectic freeway and kept staring at whatever was crossing the overpass, you probably would be getting into numerous roadway accidents down below.
Okay, you are probably thinking, it might seem unlikely to notice the dump truck when it was properly and without noticeability going across the overpass. But what about once it struck the railing, certainly you would notice it then. At that point, your concern would rise like that of a startled cat that gets poised for action.
Being a catlike driver, you would then be eyeing the dump truck and waiting to see what happens next.
Maybe the whole darned truck is going to fall onto the freeway. Perhaps broken-off parts of the truck and bits of railing will come down. And so on.
Would you really though be thinking that hazardous asphalt is going to massively drench down upon the freeway?
Seems nearly unimaginable.
It is also doubtful that you would realize that the dump truck was hauling asphalt. You would need to be very observant to not only spy the dump truck, see that it is hitting the railing, watch as the truck pivots, and also keenly observe that it so happens to be carrying asphalt. That’s a lot of unbelievably astute observations. Pat yourself on the back.
The thing is, the entire incident took place in just a few split seconds. Entirely out of the blue. You weren’t expecting it to happen. I’ll therefore ask the question again, doing so in the context that you are caught entirely flatfooted, not necessarily paying attention to the overpass, and are primarily and sensibly worrying about the fast-flowing traffic surrounding your car on an everyday freeway.
Would you anticipate that asphalt is going to flow down upon the freeway?
Realistically, the answer is no.
Shifting gears, the future of cars consists of AI-based true self-driving cars.
Allow me a moment to elaborate.
There isn’t a human driver involved in a true self-driving car. Keep in mind that true self-driving cars are driven via an AI driving system. There isn’t a need for a human driver at the wheel, and nor is there a provision for a human to drive the vehicle. For my extensive and ongoing coverage of Autonomous Vehicles (AVs) and especially self-driving cars, see the link here.
Here’s an intriguing question that is worth pondering: How would an AI-based true self-driving car do in the extraordinary circumstance of the dump truck that has spilled hot asphalt from an overpass onto an active freeway?
Quite an interesting puzzler.
Before jumping into the details of what would likely happen, I’d like to further clarify what is meant when I refer to 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 Falling Flying Asphalt
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.
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 consideration in this dump truck scenario entails the potential detection of the showering asphalt.
As earlier pointed out, we would not have expected that human drivers would necessarily be intently observing the overpass when everything appeared to be normal and hunky-dory. Only if something suspiciously stood out would a human driver take notice of the traffic above them on the overpass.
Once the dump truck struck the railing, this unseemly act would be undoubtedly sufficient to arouse a semblance of attention by human drivers on the freeway. Likewise, as the dump truck pivoted and the fluid-like asphalt poured out, you can bet that this would be overtly noticed. Of course, upon the asphalt cascading onto the freeway itself, there is no longer much debate that this would absolutely be seen by the human drivers (it is nearly inconceivable otherwise, as per the famous quip in The Princess Bride).
Consider the various stages of observable steps in the dump truck dumping event: A normal passage of a truck on an overpass, followed by the railing crash, then the rapid pivoting of the truck and spewing of the asphalt, finished by gobs of asphalt raining in the air and thudding onto the freeway.
What would take place if an AI self-driving car was on that freeway and in the midst of approaching the eventual asphalt apocalypse?
Detection is vital to all of this.
A self-driving car is typically equipped with a sensor suite to try and identify the driving scene. The sensors might include video cameras, radar, LIDAR, ultrasonic units, thermal imagining, and other such sensory devices. Those devices are mounted on the self-driving car. During a driving journey, the sensors are switched on and avidly collecting data. The data is being fed into the AI driving system.
The AI driving system is usually composed of Machine Learning (ML) and Deep Learning (DL) algorithms that computationally are doing pattern matching. This pattern matching is trying to analyze the data to discern what kinds of objects exist nearby and what those objects are. Along with assessing the nature of the objects, there is also a mathematical effort taking place by the computers to figure out the speed of those objects and their respective direction or headings.
We can reasonably assume that a well-equipped self-driving car would be ascertaining the traffic on the freeway that was in front of the autonomous vehicle, along with traffic adjacent to and also that is directly behind the self-driving car. The video camera would be streaming in the visual scene. The radar would be pinging off the objects that are nearby. Etc.
The AI driving system ought to have been programmed to detect the overpass if nothing else to gauge the allowed height underneath.
Any kind of bridge or structure that extends over the freeway is deserving of computational analysis to try and ensure that the self-driving car will be able to pass underneath it safely.
Imagine that a bridge had collapsed onto the freeway or at least was bulging downward, thus narrowing the gap between the freeway and the ceiling level of the structure. We would certainly expect a human driver to notice that there is insufficient clearance and therefore would stop before striking the low-hanging elements. A well-designed and properly programmed AI-based self-driving car should be doing the same.
When I mention this caveat that the self-driving car ought to be doing that kind of analysis, I am trying to purposely indicate that you don’t know for sure that an AI driving system has been programmed with that kind of capability. It could easily be the case that an AI driving system does not have that feature and the presumed thinking by the AI developers is that it is a so-called edge or corner case.
An edge or corner case is considered any aspect that seemingly is outside the core of the primary task at hand. If some automaker or self-driving tech firm decides that low-hanging bridges are an edge or corner case, they would put this consideration on a list of low priorities. At some future time, they would come back to that item and opt to program it into the AI driving system.
You would naturally assume that something as fundamental as discerning whether there is sufficient clearance for the self-driving car would axiomatically be included in all AI driving systems. Generally, you might be right, but there is no ironclad basis to stake a claim that this is for sure something as yet programmed. By and large, most of the AI driving systems have this, though some might not (or, they have a crude version and for which numerous exceptions would be missed or unaccounted for).
To simplify this discussion, let’s assume that we are envisioning an AI driving system that is especially robust and has all of today’s presumed everyday capabilities.
Start the clock in terms of the unfolding event.
The dump truck enters onto the overpass, driving along perfectly fine so far.
The self-driving car on the freeway would computationally examine the streaming video of the driving scene and presumably detect the overpass. The odds are that the floor-to-ceiling gap would be estimated. This estimate would indicate that the self-driving car can easily pass underneath the overpass. The overpass does not present any issues in the sense that there aren’t any other objects protruding or anything else that makes the overpass especially notable.
Our vaunted self-driving car continues to proceed forward and the AI driving system anticipates passing underneath the overpass, easy-peasy.
The clock ticks.
Ill-fated dump truck rams into the railing.
AI driving system detects the crash. The chances are that the programming of the AI driving system would not consider the matter of any alert. This is due to the notion that nothing is protruding or extending downward that could impede or strike the self-driving car.
Clock ticks again.
The dump truck pivots to its side, asphalt begins to pour out.
The motion of the asphalt might be detected. Since the impending mass seems to be heading downward, there is an off-chance that this might be calculated as a potential future issue. The prediction component in the planning portion of the AI driving software calculates the distance of the self-driving car to the overpass, the speed of the self-driving car, the speed of the falling asphalt.
A potential intersection of the two is calculated.
It is highly unlikely that the AI driving system computational pattern matching is able to identify what the falling mass consists of. Humans would likely recognize it as asphalt, due to the color, shape, movement, etc., and by logically deducing that it is material coming directly from the bed of a dump truck.
That is pretty much a form of common-sense reasoning for a human being.
Humans know about the world around them. They can figure out things by tapping into their semblance of common sense. AI systems do not yet have any kind of common sense. The development of common sense for AI is an unsolved problem that has continued to stubbornly defy numerous Herculean efforts to devise (see my coverage at this link here).
In short, unless the AI developers had somehow envisioned that someday the AI driving system needed to do pattern matching on fluid-like asphalt (put that into the no-chance bucket), the mass of falling stuff is classified by the AI as an unknown.
Whatever this falling stuff is, it could be dangerous. On the other hand, it could be soft ash and fall to the freeway with no significant weighty impact (though perhaps obstructing vision, which is bad in its own right). The AI driving system would mark this as undetermined.
Make another click of the clock.
Wham, the asphalt lands onto the freeway.
This is now totally within the usual purview of the AI driving system. The self-driving car sensors are tuned to find obstructions and debris that sit straight ahead of the autonomous vehicle. That’s pretty much a given.
What should the AI self-driving car do at this juncture?
The odds are that the AI driving system will attempt to stop the autonomous vehicle before it strikes the mysterious material. Though the self-driving car might also be programmed to consider changing lanes, given the narrow time window of this event as it unraveled, the odds are that a lane change was summarily dismissed as an immediately viable option, if calculated at all.
Conclusion
Anyone taking bets on what today’s self-driving cars might do in this harrowing circumstance would have to go with the notion that the detection would occur primarily once the material landed or was somewhat mid-air but directly in front of the autonomous vehicle. The resulting AI programming would likely go with the seemingly default response of halting the vehicle.
It would be surprising to envision that any of today’s AI driving systems would do much more than that.
That is not to suggest that future iterations and more advanced versions would similarly be limited in what they detected and what options were calculated and computationally assessed. Ongoing efforts to further expand the capabilities are stridently occurring.
There is a bit of the nefarious Trolley Problem that sits at the ugly underbelly of this dump truck scenario.
The Trolley Problem entails having to make difficult choices, contextually herein in the nature of driving decisions, and is often offhand dismissed by some as a misleading folly in the self-driving realm. In contrast and as a stout contrarian on the matter, I instead argue ardently that the Trolley Problem is a useful tool and applies to the real-world and practical use of self-driving cars, see my extensive discussion at this link here.
Overall, a claimed advantage for self-driving cars will be that if an AI driving system encounters an extraordinary event, this can then be used by the AI developers as a learning insight that can be programmed into subsequent versions of an AI driving system.
Go ahead and thusly add this: Dump truck on overpass dumps a load of noxious asphalt onto freeway traffic below.
Hopefully, we can add that to the school-of-hard-knocks catalog of lessons learned.
Post a Comment