Slushy sludgy gunky snow is shoved up to the curb and becomes either a hazard or a help toward … [+]
Boston Globe via Getty Images
Driving in the snow can be quite troubling.
If the snow is coming down while you are underway on a street or highway, you might not be able to see clearly and therefore could misjudge a driving situation. The next thing you know, bam, you’ve rammed into another car. Or you swerve across lanes of traffic because you are frustratingly unsure of where your lane is and where the adjacent lanes are.
Those are the dangers associated with falling snow.
Of course, once the snow plops onto the roadway surface, additional problems rear their ugly heads. Your car can get readily stuck in a bank of snow. Equally disturbing, the snow might cause the roads to become slick and unreliable for traction. Other drivers can easily make inadvertent mistakes and slide like one of those wayward ice rink skaters right into your car.
For my coverage about issues associated with driving in the snow, including and especially for AI-based self-driving cars, see the link here.
All in all, there are plenty of reasons why driving in the snow can cause you great anxiety and potentially produce unseemly untoward results. One supposes that the better use of snow would be to make a snowman or perhaps have a playful snowball fight. The thing is, if you need to get yourself from point A to point B, sometimes a car is the only viable means and you have to risk driving in the snow, like it or not.
I have another aspect about snow that you might not have realized is also a bit of a problem. This is something that you’ve undoubtedly noticed and even experienced, yet you probably didn’t put much thought into the matter at hand.
Are you ready for a surprise?
Another snow-related driving consideration entails the snow and ice that end up at the curb and sit there blocking part of the roadway. The snow is at times shoveled off the sidewalk and down into the gutter next to the curb, while meanwhile the snow is pushed or snowplowed out of the mainstay of the roadway and smushed up against the curb. Bottom-line is that the snow sits in a veritable no man’s zone, no longer bothering particularly the pedestrians walking on the sidewalk, and nor seemingly bothering the cars that are trying to seamlessly make their way down the roadway (the snow is now pushed aside, by and large).
But I don’t want to leave the impression that the accumulated snow and ice that resides in that gutter setting is welcomed and a treasure.
Far from it.
We’ll start with the viewpoint of pedestrians and make our way to the perspective of drivers.
A pedestrian that wants to cross the street at a marked intersection is bound to have to cope with a mound of this often dirty and ugly carcass of snow and ice. Imagine yourself standing at the curb and looking down at the less than picturesque snow and ice abutting the curb (this is probably intermixed with soot, discarded trash, oil from leaking cars, etc.).
Do you plant your feet directly into the semi-frozen morass or do you try to scoot over the top of it? Will you fall down as you attempt to navigate across it? Should you step around the snowy blockage, but if you do so, maybe you come closer to the crazy traffic that is already scarily contending with those death-defying roadway conditions.
Decisions, decisions, decisions.
What about the drivers?
They presumably observe that the accumulated snow and ice are jutting out from the curb. One hopes they notice this. Most sane drivers try to avoid the snowy junk. No sense in getting your tires enmeshed into something that you aren’t sure what it might oddly contain (maybe nails, or other protruding and tire popping contrivances). Plus, you can lose the available traction that you have on the roadway surface that is presumably clearer and less snow and ice-covered than what you see at the curbside.
Much of the time, you need to take a much wider turn at a corner due to the accumulated snow and ice that sits adjoining the curb. You might also need to give greater clearance as you approach a straightaway curb at any point, such as when trying to drop someone off or picking up a friend or passenger.
I would dare say that drivers accustomed to driving in snow-related circumstances are used to those mounds of snow and ice that reside abutting the curb. You probably do not give much mental attention to the phenomena. It is what it is. You see the stuff, you avoid it if you can.
I’ll add an additional twist to the matter.
When you give this otherwise ordinary and dreary subject matter a careful modicum of mental noodling, you’d realize that the mounds of snow and ice are essentially narrowing the available roadway in terms of the available passageway for cars and other vehicles.
Yes, the snow and ice take up a quite otherwise considered limited resource, namely the width of the roadway that you are driving on. This might not seem especially obvious at the straightaway parts of a street or avenue. On the other hand, when you look at the corners such as at an intersection, you become more aware of how the roadway space is being constricted.
Those tight turns at the corners are no longer particularly feasible. You have to make a widened turn.
This might seem simple and without any concern. One problem is that by making a wider turn, you can perilously disrupt other traffic. Perhaps you make a right turn on red, and meanwhile, other traffic is flowing in the lane adjacent (left of) the lane you are turning into. Normally, the turn would be fine, since you would stay close to the curb and seek to not intrude into the neighboring lane.
Unfortunately, you now take the turn more widely and threaten innocent traffic in that other lane. The other traffic might have to abruptly halt or swerve to avoid your sloppily turning car. Plenty of such roadway collisions occurs as a result of the attempt to avoid the piled-up snow and ice that sits at the curb corners.
I’ll even add a new word into your vocabulary (I’d bet that most people don’t know this word).
Sneckdown.
This oddball made-up word “sneckdown” has generally become known as an apt descriptor for the snow and ice that blocks along the curb. Said to be a mashup of the words snowy and neckdown, the catchphrase or catchword was reportedly popularized via Twitter in 2014. There is ongoing competition such that some prefer to use other made-up contractions like snowspace, plowza, snovered, slushdown, and other assorted monikers.
For those devoted to transportation and the study of traffic patterns, a sneckdown can be construed as a form of so-called traffic-calming. You are indubitably familiar with traffic-calming by having had to deal with speed bumps. A speed bump is usually intended to slow down traffic, making the roadway presumably safer for drivers and also potentially safer for pedestrians too.
You could say that the weather gods provide us with a form of traffic calming by dumping snow upon us. We tend to go more slowly and drive more cautiously. That’s good for safer driving, and good for a safer crossing of the street by pedestrians.
We know though that life rarely provides a free lunch.
The snow giveth and it taketh away in terms of safety aspects. As just mentioned, the sneckdown might worsen driving by forcing drivers to take wider turns. You can look at this as a glass-half-full that the wider turn is hopefully a slower turn and less likely to bump into any pedestrians that are standing at a curb. The glass is also half-empty because you can misjudge the wider turn and end up disrupting other traffic or otherwise making for a traffic uncalming experience.
Anyway, believe it or not, transportation planners often study the effects of sneckdowns to try and ascertain whether perhaps something equivalent should be permanently put in place. Thus, rather than waiting for the snow to come along and possibly make the corners safer, the idea is that you might repave or change the markings on the roadway so that there is always a kind of curb extension at hand, including during winter and summer weather.
You also need to be worried about other exigencies. For example, will emergency vehicles such as speeding ambulances or fire trucks be better off or worse off due to some kind of sneckdown that is permanently established? You can make the case for the good and the bad, depending upon a particular curb setting and factors including the nature of the street, the width of the street, the volume of traffic, the volume of pedestrian activity, etc.
No easy answers on this one.
Shifting gears, we’ve so far discussed that human drivers can be troubled with sneckdowns. I’d like to introduce a different angle on the sneckdown topic, namely the advent of AI-based true self-driving cars.
Here’s a noteworthy question that is worth pondering: How will AI-based true self-driving cars contend with those exasperating snow-based nature-produced (kind of) sneckdowns?
Allow me a moment to unpack the question.
First, note that 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, 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.
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 The Sneckdown Predicament
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 is 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.
First, it is important to realize that not all AI self-driving cars are the same. Each automaker and self-driving tech firm is taking its approach to devising self-driving cars. As such, it is difficult to make sweeping statements about what AI driving systems will do or not do.
Furthermore, whenever stating that an AI driving system doesn’t do some particular thing, this can, later on, be overtaken by developers that in fact program the computer to do that very thing. Step by step, AI driving systems are being gradually improved and extended. An existing limitation today might no longer exist in a future iteration or version of the system.
I trust that provides a sufficient litany of caveats to underlie what I am about to relate.
We are primed now to do a deep dive into how self-driving cars might cope with sneckdowns.
The best place to start consists of how most AI self-driving cars currently navigate our roadways. Whereas you might assume that the AI driving system is encountering a roadway for the first time upon initially perchance coming upon that street, this is not usually the case.
You see, the norm these days consists of first extensively mapping whatever roadway system a self-driving car is going to be used within. An automaker or self-driving tech firm might declare that they are going to deploy their self-driving cars in a particular 10-block wide area or within the boundaries of a specific city realm. Before they put their self-driving cars onto those streets, they usually do a rather exhaustive mapping exercise.
They might use conventional human-driven cars to drive throughout the designated area. Those cars are outfitted with special mapping technology. The digitally captured maps are then collated, processed, and refined, being further enhanced and reviewed.
They might then use their self-driving cars, albeit with a human backup or safety driver at the wheel, and do even more residual or last-mile (last-feet) type of mapping. At some point, the mappings are put into use and then tweaked or augmented as required.
Normally, wherever you see an AI-based self-driving car rolling along, the odds are pretty high that a huge pre-mapping activity occurred long before that self-driving car did so while absent a human backup driver at the wheel.
I realize that some of the mapping vendors will scream to the rooftops and exhort that they sell digital maps and there is little or no need for automakers or self-driving tech firms to create their own. Those vendors would decry that there is no sense in reinventing the wheel, so to speak, and instead to use their readily available and detailed maps.
To some degree, they are right, namely that in some instances it makes the best sense to simply purchase or license such digital maps. In other cases, the automaker or self-driving tech firm might have other bona fide reasons to craft their own set of maps. There are tradeoffs involved in which way to go (I’ve covered this in prior columns).
Why have I dragged you through this seeming minutia about the use of digital maps for self-driving cars?
Because the sneckdowns can be a digital map’s worst nightmare.
Let’s assume that it is a nice summer day. A self-driving car is going to make the turn at an upcoming corner. Due to the detailed maps of the area, the precise location of the curb is already known, along with the distances to all other nearby pertinent points. The AI driving system can calculate with rather high precision a suitable approach to making that turn, encompassing a calculated distance to avoid striking the curb and yet a close enough distance to not unduly widely take the turn.
This is somewhat easy-peasy.
We can complicate matters by having a pedestrian stand at the curb and start to step into the street. The AI driving system via the onboard sensors such as a video camera, LIDAR, radar, and the like would hopefully detect the intrusion by the pedestrian. This then would get factored into how to safely proceed to make the turn.
All of that is the straightforward aspect of programming an AI driving system to deal with such things as making turns at corners. By having pre-mapped the settings, the only surprises are whatever perchance happens at the moment, otherwise, the rest of the driving scene is relatively known and already ascertained.
Let’s change the scenario and dump some of that sneckdown into this glorious driving scene.
An AI self-driving car is coming up to make the turn at the corner. The digital map says that the curb is at geographic point Z. By normal calculations, the turn can be made within a few inches of that position.
If the AI driving system via its sensors does not detect that the sneckdown is sitting there, the self-driving car can be guided into making a turn that won’t work out very well. The autonomous vehicle might steer right into that patch of clumpy dirty snow and ice. The next thing you know, oops, the self-driving car is going awry of the pristine path that was plotted.
All heck breaks loose.
Nobody wants that to happen.
Okay, so the AI driving system has to be programmed sufficiently to deal with sneckdowns. This means that the sensors have to detect the sneckdown, including how large it is, the likely viscosity of the morass, and so on. Human drivers generally do this rather easily (not all humans, that’s for darned sure!), but getting the AI driving system to do this with significant aplomb is a lot harder than you might think.
As you can see, a predominant reliance on a digital map that was made during summer weather that lacks sneckdowns can be a far cry from what the roadway is like during the winter snowy months. You can either try to create alternative maps, or possibly send out those mapping vehicles to remap throughout the inclement weather periods, or just deal with the fact that during certain months of the year those sneckdowns will magically appear and the AI has to be programmed accordingly to compensate for the matter.
There are already a lot of vital aspects for the AI driving system to have to be contending with, especially when the weather gets snowy and icy. Chewing up tons of computer processing cycles to calculate and recalculate that those sneckdowns can be computationally problematic. And, this presumably needs to be done upon each encounter with a sneckdown.
Some of the automakers and self-driving tech firms attempt to provide a global update to their fleets upon discovering semi-permanent changes in the roadways. In short, suppose a self-driving car comes randomly upon a rather tricky sneckdown.
Luckily, let’s assume the AI driving system handles it well.
During some later time of day that the AI driving system uses its OTA (Over-The-Air) electronic uploading and downloading capabilities, it might push up the fleet cloud a noted alert or indication about the discovered sneckdown. Now that this is reported into the cloud, the developers could analyze the sneckdown or might have an automated capability to do so. This in turn can be used to produce some newly figured out driving guidance for the AI driving systems as to how to deal with that particular sneckdown (just in case any other of the autonomous vehicles in the fleet perchance come upon that corner and that specific sneckdown).
Via the OTA, updates to the AI can be pushed down into the self-driving cars from the fleet cloud, such as the latest software patches, and included could be the new guidance about that sneckdown at location Z, and here’s how to prudently traverse it. In this manner, the whole fleet can now be ready to deal with the sneckdown and navigate accordingly.
I don’t want that to seem overly easy, and please know that a slew of complexities can make this a lot harder. I’m just outlining some of the rosier highlights. There are plenty of lowlights, if you will, whereby things can go wrong or make for added troubles.
Conclusion
In the field of self-driving cars, there is something known as the ODD which stands for the Operational Design Domain (see my detailed explanation at this link here). Automakers and self-driving tech firms will indicate what kind of driving situations their existing AI driving systems can properly handle, doing so is considered defining the ODD for those particular autonomous vehicles.
For example, we might define one kind of ODD as consisting of having an AI driving system that can work in relatively sunny weather, no snow or ice, mild rain is okay but not heavy rain, and all of this is within the bounds of a five-mile pre-mapped area of some designated city R.
The more you pile harsher driving conditions into an ODD, the tougher it is to craft the AI driving system to deal with those less manageable driving situations. This is partially why many of the self-driving cars were first being tested and rolled out in locales that have milder weather or simpler roads (like certain parts of Arizona). When you rachet up to snowy and icy weather and chaotic or byzantine roads and wild traffic, the hurdle of developing sufficient AI driving systems gets a lot higher.
You can add sneckdowns into the thornier classes of ODD.
This might seem surprising to those of you that routinely cope with sneckdowns.
Something that human drivers seem to usually take in stride can in fact be a daunting challenge for AI driving systems. I realize that some pundits might argue that a sneckdown is merely an edge case or considered an outlier related to the kinds of driving issues that an AI self-driving car has to contend with. Edge cases are those driving aspects that are ranked as low priority and can be dealt with later on, after first having dealt with the considered higher priority or core aspects of doing day-to-day driving.
You can toss the sneckdown into the low priority heap if you wish, though I assure you that when people see an AI self-driving struggling to deal with a sneckdown, and (heaven forbid) the AI driving system miscalculates and troubles arise, the sneckdown importance might become more important than originally conceived.
As a daily human driver, I give due respect to sneckdowns. My hope is that if I treat them right, they will treat me right in return. When you next get a chance, go ahead and calmly give a tip of your hat to the lowly, oft-ignored, and yet proud sneckdowns that line the snowy and icy streets of your community.
They deserve a bit of sunny good wishes, doing so before they melt away.
