By Lance Eliot, the AI Trends Insider
Those driving robots.
One approach to achieving a self-driving or driverless car would be to make a robot that could drive a conventional human-driven car in place of the human. Thus, rather than building the driving capabilities directly into a car, the robot would be able to get into a normal car and drive. This kind of exploratory activity is taking place in some research labs, but generally is a long way afield of being practical.
In the near term, the only viable route to a true self-driving car is going to involve infusing the self-driving tech into the vehicle itself. But just because something is not immediately or near-term achievable doesn’t mean that it should not be pursued. Those having a longer-term time horizon can envision a future that would have robots for driving cars, trucks, buses, and so on.
For those researchers pursuing the driving robot, keep up the noble work and do not be deterred in your valiant pursuits. As they say, keep on truckin’.
It might seem that it wouldn’t make a whit of difference whether the car drove itself via built-in capabilities or whether a robot got into the vehicle and did the driving. In my columns, I have identified several key benefits for having a robot that could drive a conventional human-driven car:
Any conventional car could instantly become a self-driving car, albeit being driven by a robot, but at least not requiring a human driver, and do so by simply having the driving robot get into the vehicle and perform the entirety of the driving act.
You could presumably easily switch the driving robot from one conventional car to another. This means that any car at any time can immediately become a self-driving car (i.e., non-human driven).
The driver of the car would be readily seen. This is a somewhat less obvious downside about the upcoming self-driving cars, namely that with the built-in driving capabilities there isn’t a driver in the driver’s seat. Worries abound that when human drivers in other cars attempt to look at the driver of a self-driving car, they will not have any cues about the driving acts, and likewise, pedestrians will lack such cues too. Various add-ons might be needed for self-driving cars to overcome this qualm.
The cost of having a self-driving car goes toward the driving robot rather than the car itself. In brief, one concern about true self-driving cars is that they might be exceedingly pricey, perhaps becoming something only the wealthy can afford. A driving robot would drive an everyday car, meaning that the car itself is still the same price regardless of whether the robot or a human is driving it.
A driving robot could be more readily shared around and used on a widespread basis. A true self-driving car with built-in capabilities is merely one car. A driving robot could drive any conventional car. As such, the driving robot has greater utility, plus the cost of the driving robot can be spread amongst a multitude of users or owners in a more versatile way than could a singular self-driving car.
A driving robot might provide additional uses. A true self-driving car has just one purpose, ostensibly it is a car that drives and that is all that it does (though, notably, this is a darned impressive act!). A driving robot might be able to perform other tasks, such as being able to get out of the car and carry a delivery package to the door of a house. Note that this is not a requirement per se and merely identifiable as potential added use that might be devised.
There are also various disadvantages of using a driving robot versus aiming to utilize or craft a true self-driving car, which I won’t delineate those shortcomings here. I urge you to take a look at my earlier article on the topic to see the articulated list of downsides or drawbacks. The only reason I am not listing the disadvantages herein is merely due to the amount of ink space that it would consume and want to instead focus on some additional elements on the topic of driving robots.
Meanwhile, please do know that there are disadvantages to driving robots and I trust that you are aware that a driving robot is not all roses and can also contain prickly thorns.
One other quick point. Do not fall into the classic mental trap of thinking that these two approaches have to be mutually exclusive of each other. That’s a false dichotomy. We can have true self-driving cars that have all the AI guts built-in directly into the vehicle, and we can also have driving robots.
This might seem puzzling. Wouldn’t we only need one or the other, and not need both?
In the United States alone there are about 250 million conventional cars. Some pundits are dreamers that espouse that somehow, magically, overnight all of those conventional cars will disappear or evaporate into thin air, once we have viable self-driving cars. This doesn’t make economic sense and the reality is that it will be many years, likely decades before we gradually see a predominant prevalence of self-driving cars and the winnowing of conventional cars on our roadways.
As such, it makes perfectly good sense that you might have some portion of cars on the highways and byways that are true self-driving cars, and simultaneously have conventional cars that are being driven by driving robots. Doing both such approaches is absolutely sensible and certainly doable, assuming that self-driving cars are achievable and those driving robots are achievable.
One final comment before getting into the meat of this discussion and covering more details about driving robots, none of this says anything about forcing human drivers out of the driver’s seat. All told, you could be on any given freeway, highway, or street, and witness next to you a true self-driving car, and a driving robot in a conventional car, and a human-driven car nearby too.
They would all co-exist on our roadways.
This is a bit of a disappointment for some that want to excise or make extinct the act of human driving. Their sentiment is certainly in the right place due to the hope of forever doing away with drunk driving, distracted driving, and the other foibles of human drivers. The U.S. has about 40,000 car crash fatalities annually and about 2.5 million injuries due to car crashes, of which much is due to human driving shortcomings. In that frame of reference, the less there is of human driving, and assuming that true self-driving cars and robot drivers are as safe or safer than human driving, the better off we all are.
A sticking point is whether doing away with human driving is a societally feasible option remains an open question. Essentially, even if you could scale-up the self-driving cars and the driving robots, thus imagining that no human driving was needed, would we indeed forbid or ban all human driving on our public roadways?
Right now, some people would argue stridently that the only way you’ll stop them from driving is when you remove their cold dead hands from the steering wheel.
Shifting gears, let’s get back to those driving robots.
Here is today’s intriguing focus: What kind of ground rules ought there be about those driving robots?
Let’s unpack the matter and see.
For my framework about AI autonomous cars, see the link here: https://aitrends.com/ai-insider/framework-ai-self-driving-driverless-cars-big-picture/
Why this is a moonshot effort, see my explanation here: https://aitrends.com/ai-insider/self-driving-car-mother-ai-projects-moonshot/
For more about the levels as a type of Richter scale, see my discussion here: https://aitrends.com/ai-insider/richter-scale-levels-self-driving-cars/
For the argument about bifurcating the levels, see my explanation here: https://aitrends.com/ai-insider/reframing-ai-levels-for-self-driving-cars-bifurcation-of-autonomy/
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 a Level 4 and Level 5, while a car that requires a human driver to co-share the driving effort is usually considered at a 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).
Since semi-autonomous cars require a human driver, the adoption of those types of cars won’t be markedly different from 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.
For why remote piloting or operating of self-driving cars is generally eschewed, see my explanation here: https://aitrends.com/ai-insider/remote-piloting-is-a-self-driving-car-crutch/
To be wary of fake news about self-driving cars, see my tips here: https://aitrends.com/ai-insider/ai-fake-news-about-self-driving-cars/
The ethical implications of AI driving systems are significant, see my indication here: https://aitrends.com/selfdrivingcars/ethically-ambiguous-self-driving-cars/
Be aware of the pitfalls of normalization of deviance when it comes to self-driving cars, here’s my call to arms: https://aitrends.com/ai-insider/normalization-of-deviance-endangers-ai-self-driving-cars/
Cars And Those Driving Robots
I’d like to identify some ground rules about driving robots.
Upon making such a bold statement, let’s be abundantly clear that a driving robot can be crafted in whatever way anyone wants to do so. Assuming that the driving robot is able to perform the task of driving, and furthermore assuming it abides with any rules or lawful regulations, the design, and other facets can be whatever you want.
As an aside, you can bet your bottom dollar that if we do reach a point of having driving robots, this will raise all sorts of interesting questions about how to ensure they are proficient and properly-being permitted to drive a car. Just as humans have to undergo various requirements to get a valid driver’s license, so too would those precious driving robots.
Envision, if you will, a driving robot that is standing in line at your local DMV (Department of Motor Vehicles), waiting to take the driver’s test.
Of course, that’s a bit of a ridiculous form of theater and not likely how the regulating and certifying of driving robots would occur. There would undoubtedly be all sorts of vital checks-and-balances underlying the deployment and use of driving robots. We certainly don’t want to end-up replacing human drivers with something worse that causes more harm than good.
Anyway, if you mull over the notion of a driving robot, some handy ground rules might come to mind. Rather than having you start that list from scratch, I’ve put together my own list that can provide fodder for discussion on the topic.
Each of the ground rules is “reasonable” in the sense that there is a rational basis for proposing the ground rule. I can only cover so much of the underlying logical basis herein, plus I’ll try to briefly explain what makes the ground rule worthwhile for consideration. I’m sure you’ll come up with even more ground rules to be added to the list, and in which case, that’s wonderful, and the more the merrier.
Here you go.
- The driving robot must be mobile in its own right.
This ground rule is that the robot must be able to enter into and exit from the car, doing so on its own, without any human assistance. If it required a human to get it into place, my qualm is that this would severely limit the utility of the thing. Just as a human can usually get into and out of a car on their own, so should the driving robot. This also allows then for a human to readily take over the driving, when so desired, by simply asking the robot to get out of the way. It also means that a human can presumably tell the robot to get into the car and start driving. No need to deal with picking up the beast or otherwise lifting a finger to get it to become the car driver.
- The driving robot must be able to fit within the normal driver’s space.
For this ground rule, the notion is that the robot cannot be larger than the usual amount of space set aside in a car for a driver to sit and drive the car. If the robot was oversized, it would then not be able to drive in most cars. Furthermore, suppose it took up the space of say two people, well, there aren’t many cars these days that would have a front bench-style seat to accommodate this. The driving robot has to be able to sit in a normal-sized driver’s bucket seat and cannot use up any space beyond that of a rotund human.
- The driving robot must be able to drive the car entirely on its own.
In this ground rule, the point is that the driving robot cannot “cheat” and somehow require someone or something else to aid it in doing the driving. For example, imagine if a human needed to do the steering, meanwhile the driving robot did everything else. I reject that idea. The driving robot has to be good enough that no other adult driver is needed in the car. This means that the driving robot can completely undertake the driving task when there isn’t an adult present, and indeed can drive on its own when no human passengers are on-board either.
- The driving robot must use the existing human-based driving controls.
This ground rule is that the driving robot must be able to use the conventional steering wheel, brake pedal, and accelerator pedal, along with being able to access and use all the other conventional operational driving elements in the car. You might be tempted to suggest that the driving robot could merely plug into some specialized port of the car, and issue electronic driving commands. Yes, that’s fine as a bonus, but this also assumes that a conventional car has such a provision. Almost none do, and you would need to convince the automakers to start changing their cars to allow for this possibility. Nope, the car is a car, and the driving robot has to drive it as it exists, just as humans drive the car. That is a core requirement.
- The driving robot can plug into the car for power, but only conventionally so.
I am going to somewhat carve out a kind of exception to the prior ground rule, allowing for the driving robot to potentially plug-in and draw electrical power from the car. But don’t get too excited about this. The restriction is that the driving robot can only plug-in and get power in the same manner that you could plug-in any other electrical accessory. In that sense, the car doesn’t need to be outfitted with anything special to provide power for the driving robot (it will undoubtedly need its own source of power). In short, this willingness to have the driving robot use the power of the car is not that big a deal, since it is restricted to what one can do with any car already, and likely will not especially solve any power consumption design problems for those devising a driving robot. This won’t make things much easier in that regard, sorry.
- The driving robot cannot depend upon any kind of ADAS (advanced driver assistance system) that exists for the car.
This ground rule is similar to the earlier rule that the driving robot must be able to drive the car on its own. Some cars have almost no semblance of ADAS, and as such, the driving robot has to be able to drive those kinds of cars. That being said, if a car does have various ADAS, the driving robot is welcome to use the ADAS, similar to how any human driver might do so. Keep in mind that the driving robot is still responsible for the driving (in a technological sense, not necessarily a “legal” meaning), and it cannot rely on the ADAS any more than a human might. The ADAS cannot become a needed crutch for the driving robot to be able to undertake the driving task.
- The driving robot must be able to interact with passengers, doing so normally.
This is a somewhat trickier ground rule. For a true self-driving car, the automakers and self-driving tech firms are still figuring out how to enable the passengers to interact with the AI driving system. In some instances, you convey your destination via say a smartphone, and that’s about it. The future is supposed to entail having the AI driving system using Natural Language Processing (NLP), thus the passenger can converse with the AI driving system to indicate preferences like swinging through a drive-thru burger joint amid a driving journey. For a driving robot, there have to be some viable means to have passengers communicate with it, else the driving robot would be a rather stilted driver and the driving act would be a lot less valuable.
- The driving robot must be reliable, safe, secure, trustworthy, etc.
This ground rule is the usual kind of table stakes for any robotic system that might be interacting with humans and that is performing a truly life-or-death type of task. If the driving robot sometimes hiccups and is going to freeze-up with a blue screen of death, well, that’s certainly not the kind of driving system anyone would want. Any and all driving robots that are going to be at the wheel of a car must be reliable, safe, secure, trustworthy, etc., and imbue a slew of such vital and absolutely required traits. Unlike humans that might decide to take a drive when they are out-of-touch or out of their gourd, the driving robot must not drive if it isn’t fully ready, capable, and able to drive. Enough said.
- The driving robot can do bonus aspects but must at least do the minimum set.
This ground rule involves the notion that the driving robot must be able to do all the normal activities that a human driver would or could do as it relates to the driving task. That is the bare minimum. On top of that, additional icing can be placed on top of this cake. Can a driving robot also be your personal chef and make the world’s greatest souffle? Sure, but not at the expense of somehow being unable to fully and properly drive a car. The car driving chore is the forte and raison d’etre. Anything else that the driving robot can do is a nice plus, and welcomed, assuming it is civil, proper, lawful, and sensible.
For more details about ODDs, see my indication at this link here: https://www.aitrends.com/ai-insider/amalgamating-of-operational-design-domains-odds-for-ai-self-driving-cars/
On the topic of off-road self-driving cars, here’s my details elicitation: https://www.aitrends.com/ai-insider/off-roading-as-a-challenging-use-case-for-ai-autonomous-cars/
I’ve urged that there must be a Chief Safety Officer at self-driving car makers, here’s the scoop: https://www.aitrends.com/ai-insider/chief-safety-officers-needed-in-ai-the-case-of-ai-self-driving-cars/
Expect that lawsuits are going to gradually become a significant part of the self-driving car industry, see my explanatory details here: https://aitrends.com/selfdrivingcars/self-driving-car-lawsuits-bonanza-ahead/
You might have realized that the key underlying theme is that the driving robot is essentially supposed to be backward compatible with the conventions of human drivers.
In the computer field, we refer to backward compatibility as the proviso that any new hardware is fully compatible with prior hardware (that’s a quick definition). For the driving robot, the widest possible usability would be that the thing would be able to slide easily into the driver’s seat of any car that a normal human could drive, start the engine, and drive away, just like a human.
Not one iota or ounce of a change needs to be made to the car itself.
Admittedly, this is a bit of a dreamy vision. There would likely be some exceptions. The exceptions though should not become the rule. Robust and popular mechanization would ensure that the driving robot would be able to be a driver whenever you needed to have one.
For self-driving cars, you need to find a self-driving car before you can use it.
In the case of a driving robot, as long as you have a driving robot near you or accompanying you, any car will do.
You don’t need to know how to drive. You don’t need to tell the driving robot how to drive. You can entrust the driving robot to do the driving. The driving robot can drive without any instruction. The driving robot can even drive when the car is otherwise entirely empty and has no human passengers at all.
As a smarmy aside, I’m sure that someone will ask what would a driving robot do if it got into a true self-driving car?
That’s a bit of a quippy answer, it would be a passenger and do whatever any passenger might do (well, probably, some human passengers are going to try and offer driving advice to self-driving cars, so one supposes maybe a driving robot might be tempted to do the same).
Overall, if I have whetted your appetite for a driving robot, I’d suggest you band together with those that are trying to pull off a miracle and craft a driving robot and get to work with them (of the ilk described herein, which is a tall order, I assure you).
By the way, you’ll have to get in line, since I’ve already placed my order for a driving robot. Once I get mine, maybe I’ll let you borrow it, but only if you treat it right, and I get it back entirely in one drivable piece.
Copyright 2021 Dr. Lance Eliot. This content is originally posted on AI Trends.
[Ed. Note: For reader’s interested in Dr. Eliot’s ongoing business analyses about the advent of self-driving cars, see his online Forbes column: https://forbes.com/sites/lanceeliot/]