By Lance Eliot, the AI Trends Insider
I’ve ridden many times on the BART system in San Francisco (that’s the Bay Area Rapid Transit, or BART system, which provides mass transit throughout the Bay Area). Typically, during work transit times in the morning and the evening, each person is sitting (or standing) and trying to carve out their own little world while being jam packed with lots of other people in the sardine can of a train car. Many people are wearing audio headsets to block out the noises around them. Many are wearing sunglasses or something to try and mask their gaze. I’ve even recently seen some wearing a mobile Virtual Reality (VR) headset, which really does help to block out the real-world and enter into a dreamscape world.
Mass transit for many people is something they’d prefer to not be involved in. Being around lots of other people can be confining. Most people are relatively sane and allow you some privacy, but you’ve then got the wild passengers that insist on trying to draw attention. They dance, they sing, they make a spectacle. There are some that seek money. There are some that just seem to have lost their wits and are inane as they speak or make sounds or movements. It can be quite a ride.
I know it would seem like a small price to pay for being more economical about the cost of transit. I know too that it is better for the environment to have people transported via a large-scale operation. It presumably drops the pollution per capita of those traveling. It might even reduce the transportation costs per capita, per mile traveled and per person traveling. There is certainly a long list of important reasons to have mass transit.
Though the case is compelling, it goes against our seemingly innate desire to travel with control over our own space. When you are in a mass transit situation, whether being on a bus, a tram, a train, a subway, a commercial plane, etc., you are inevitably forced into being in a large group. You have no choice. Once you step into the mass transit, you become a member of the mass. You then must be around and immersed within the mass. Some commercial flights provide a bubble berth for First Class passengers that want to sleep and have their own privacy. That’s a rarity, though, and most of the time you become a member of the mass.
When you drive your car, you are in a sense “separated” from the mass. You now have your own bubble, as it were. You are within the mass, but you have your own space. You are overall separated from the mass. Of course, your car will be immersed within the mass, such as when you are on the freeway, but you still have your own space. It’s kind of amazing when you think about it – you can be bumper to bumper on the freeway, your car is a scant few inches from cars all around you, and yet you still don’t need to interact with the people in those cars, and you are able to create your own internal environment. You set the temperature as you wish, you play music as you wish, you move around in your space as you wish, etc.
One of the reasons that especially Southern California has had such a hard time of getting adoption of mass transit is that no one wants to give up their own separate space. They don’t want to give up their own control over their miniature environment in their own bubble. They want to keep their own bubble, and yet have that bubble get them over to work or to the ballgame. I realize it might seem ironic that you’d be willing to sit in a Dodger Stadium with 56,000 other unseparated people, and yet still insist on your own bubble when getting to and going away from the baseball field, but that’s what it seems to culturally be.
To be fair, for the purposes of transportation, it’s also about wanting to get to where you want to go, doing so when you want to do so. When you use mass transit, you must conform to the schedule for the masses. That BART train leaves from the Powell station at 2:10 p.m. and again at 3:05 p.m., and you need to conform to that schedule. With your own personal transportation, you decide when you want to leave. Maybe it’s 2:20 p.m., and you decide it’s time to go, well, you jump into your car and away you go.
There are mass transit pundits that will counter argue that you might indeed be able to leave when you want via your own personal transportation, but it doesn’t mean you’ll arrive when you want. In other words, the BART train might get to your desired location at 4:00 p.m. by having left at 2:10 p.m., meanwhile if you leave in your car at 2:20 p.m. you might not get to your destination until 5:00 p.m. Thus, you might be fooled or lulled into believing you are controlling your transportation, but you would actually be more efficient if you used the mass transit.
We can go back-and-forth about this. In the end, most people do not opt to use mass transit. We’ll for the moment agree that it is due to the ability to have unscheduled times of transportation, and be able to travel in small groups of your own choosing and have a controlled bubble when doing so. Whether this is selfish is another matter. Whether people are right or wrong as to what they choose to do, as based on some viewpoint of societal good, that’s again another matter too. I’m not going to get overly fixated here on a lengthy debate about the efficacy of mass transit in society.
Let’s instead discuss something related to mass transit, namely Personal Rapid Transit (PRT).
You might not be familiar with PRT. There aren’t many PRT’s in existence. There has been a fair amount of discussion and research about PRT’s. You’ve likely “seen” a PRT as portrayed in a science fiction movie or comic books. Unlikely that you’ve ridden in a real-world one.
The concept is that you would be able to leverage some aspects of mass transit, but then have some aspects of personal transit. You might view this as being the best of two worlds. There are others that would say that you actually end-up with the worst of two worlds.
Usually, a PRT consists of some kind of pods, often referred to as podcars. These podcars allow you to have your own separate space, though they usually consist of seating for maybe 3 to 6 people. We’ll call that a small group. The podcar looks somewhat like one of those airport trams you’ve probably taken, though think of it as just one train car and much smaller in size so that it seats a handful of people at most. For those of you that used to visit Disneyland in the olden days, they had the People Mover which had what might be called podcars (and it didn’t even take an E ticket to ride one!). For some ski resorts, they have a tramway that is small enough that you could consider it to be about the same as a podcar.
So, the podcar provides the first part of the equation, an ability to have your own bubble (pretty much). The second part of the equation is the unscheduled times to get to where you want to go. Let’s consider how that works.
The PRT is supposed to consist of a network of special guideways, upon which the podcars operate. The topology of the network allows for the podcars to proceed across the network and not need to stop at intermediary locations. Imagine if we had a network grid, let’s say like a spreadsheet shape, and we’ll call the columns by letters of the alphabet and the rows by numbers. You get into a podcar at position A1. You want to get to position B5 on the network grid. With a typical mass transit system, the train car at A1 might stop at A2, then A3, then A4, then A5, and then proceed to B5. You’ve made a series of stops to get to your destination. A podcar of a PRT would be intended to go from A1 to B5 without any intermediary stops. It would pass through A2, A3, A4, A5, and then arrive to B5.
Thus, you get your own space and which leaves when you want and goes directly to where you want.
How do you know that a podcar will be waiting for you at A1 when you arrive? Answer, there are either lots of podcars all throughout the network and one is always ready to go, or, instead, you bring your own podcar. Keep this in mind for a moment and I’ll further elaborate shortly.
How do you know that the podcar will go straight to B5 without stopping in-between? That’s up to the system that coordinates the movement of the podcars. It is typically a centralized control of the podcars and the system is trying to ensure you don’t stop. Of course, this is not as easy as it seems, since it depends upon the other traffic on the network and the overall capacity and availability of the network. In some cases, the podcars might also operate by peer-to-peer coordination, rather than a centralized control.
Personal Rapid Transit (PRT) Could Promote Mass Transit
The goal of a PRT is to have personal oriented rapid transit, doing so to overcome the qualms people have about using mass transit. Proponents of PRT would say that it also can be done at a lower construction cost than MRT (Mass Rapid Transit), and with less regulatory easements needed, and be less obtrusive, and so on. And would have greater acceptance by society. The PRT is built with the guideways and network being protected so that it can whisk along the podcars at very high speeds.
Opponents of PRT say it’s a boondoggle that in the end won’t approach the lesser costs of MRT and the lessened environmental advantages, and that the money spent on PRT is tossing good money after bad. The MRT proponents would say that the PRT money should be spent instead on MRT, and that PRT is an unfortunate distraction and drain away from the focus that should be on MRT.
I have no dog in this fight right now about the PRT versus MRT battle herein. If you are intrigued by the debate, go find a proponent of PRT and a proponent of MRT, and let them go at each other.
Meanwhile, here’s what I’d like to cover.
What does all of this have to do with AI self-driving cars?
At the Cybernetic AI Self-Driving Car Institute, we are developing AI systems for self-driving cars. We are also exploring innovative offshoots of AI self-driving cars. PRT is one such offshoot.
Remember that I earlier said that the podcar waiting for you at position A1 of the PRT network might be a podcar pre-arranged there for use, or it might be that you bring your own podcar. Some PRT designers have pointed out that maybe there’s not a need to have lots of standing by podcars in a PRT. Instead, if someone drivers their own car, why not turn it into a podcar, so to speak.
Here’s how it would work. You drive up on your car to a podcar station. You drive your car into a kind of shell. Your car is now inside this shell. You stop driving your car at this juncture. The wheels of the car are locked into place inside the shell. You’ve maybe done something like this when you went to a car wash. At some car washes, you drive up to the entrance, you turn off your engine, you’ve positioned the car into special wheel wells, and now a conveyor belt pulls or pushes along your car through the car wash. Same concept for the PRT, except sorry, your car isn’t going to get washed while in the PRT (but, hey, that’s another idea!).
Your car now is your personal space and it is in the shell (in essence, your car inside the shell is now in totality a podcar, for the moment). The shell moves along within the network of the PRT. When you arrive at the destination, say position B5, the shell opens up and you drive out. You’ve done the same thing likely on ferries that cross a river or body of water.
Suppose you don’t have a car? Will this PRT be only for those that happen to have a car? If so, it seems like an elitist form of transit because a person has to be able to afford a car. Answer, there will be podcars that are available in case you don’t have a car. The good news is that you presumably won’t need so many of these podcars because by-and-large most people are going to be using their cars instead.
Now, I’ll clarify, not all PRT’s are designed the same way, and thus there indeed might be a PRT that requires you must have a car and there aren’t any podcars available per se. There are other PRT’s that don’t make use of cars at all and require that you only ride in the specially prepared podcars. Just wanted to clarify.
Could an AI self-driving car ride on a PRT?
If you consider for the moment that a car simply drives into a shell and then parks, it would be hoped that any AI self-driving car could do this. There are various levels of self-driving cars. A Level 5 is a true self-driving car, which is one that is driven entirely by the AI and there isn’t any human driver involved. Cars less than a Level 5 are intended to be driven by both a human driver and the AI, doing so in a co-sharing manner. There must be a human driver present in the less than Level 5’s and the human driver must be ready to take over the car controls.
For my article about the levels of AI self-driving cars, please see:
I think we can all agree that any AI self-driving car, regardless of the level, would or certainly should be able to drive into the shell and park. This seems like an easy task, either for the AI or for the co-sharing human driver in the case of cars less than Level 5. No big deal here.
We would certainly expect too that the AI self-driving car can drive into the shell, park, and then when the shell gets whisked to its destination, the AI could drive the self-driving car out of the shell and go on its presumably merry way. Easy peasy.
One mild concern is whether the self-driving car will fit inside the shell. This is of course a question that needs to be asked about any car, even a conventional car. The shells won’t necessarily accommodate any size car. Also, a car that has ski racks on the top and looks like it is outfitted to go in a parade, it might also not fit. Generally, since most of the AI self-driving cars are based on the frame and structure of a conventional car, and even if it has a LIDAR unit on the top of the self-driving car, it would be reasonable to think it will fit into the shell.
The more interesting question is what will the AI self-driving car be doing while inside the shell and serving as the personal space ala podcar?
You could decide that the AI self-driving car is doing nothing particular special while sitting inside the shell. It is no different than a conventional car at this point. It’s just a receptacle that is parked inside the shell and (possibly) has people inside of it (note that it doesn’t necessarily have human occupants, as it could be that as an AI self-driving car it has driven itself to the PRT and it is merely wanting to use the PRT to get to some other part of town).
The AI self-driving car at this juncture inside the PRT could be considered a “dumb” car as it no longer is doing anything active. It’s simply along for the ride, as it were.
AI Self-Driving Cars Could Provide “Smarts” to PRT System
Another perspective is that maybe we use the AI smarts as part of the PRT system. Perhaps the PRT system taps into the AI of the self-driving cars that are in the PRT, and uses them to help coordinate the movement of the podcars. In fact, since AI self-driving cars are intended to likely have V2V (vehicle to vehicle communication), it could be that the AI self-driving cars all operate within the PRT to do a peer-to-peer coordination of the movement of the podcars.
This then leverages the otherwise dormant AI and computer processing that is sitting idle while the AI self-driving car is acting like a passenger in the PRT. Any AI self-driving car that wanted to ride the PRT would need to download an app, akin to downloading an app on your mobile phone. Once the app was installed in the AI self-driving car, it could then be an active participant in the PRT. If it did not have the app, it could not be an active participant.
One downside to this approach consists of the aspect that what about conventional cars that might want to also use the PRT? They don’t have the computer processing and the AI to actively participate in the PRT transportation coordination. You could say that those conventional cars, or those AI self-driving cars that don’t decide to actively participate, are then coordinated by the AI self-driving cars that are participating. Or, maybe a centralized control system then deals with those podcars. It is something that needs to be considered and figured out.
For my article about swarm intelligence and AI self-driving cars, see: https://aitrends.com/selfdrivingcars/swarm-intelligence-ai-self-driving-cars-stigmergy-boids/
For my framework about AI self-driving cars, see: https://aitrends.com/selfdrivingcars/framework-ai-self-driving-driverless-cars-big-picture/
For idealism about the future of AI self-driving cars, see: https://aitrends.com/selfdrivingcars/idealism-and-ai-self-driving-cars/
The AI self-driving car could be doing other things while inside the shell, doing something other than non-driving. For most AI self-driving cars, they will need to periodically do OTA (Over The Air) updates, which will usually be undertaken while the AI self-driving car is relatively dormant. Doing the OTA during the PRT riding might be a convenient time and place, depending upon whether it makes sense to do so (if the OTA will take two hours and the ride time on the PRT is only 15 minutes, probably wise to not start the OTA).
For OTA and AI self-driving cars, see: https://aitrends.com/selfdrivingcars/air-ota-updating-ai-self-driving-cars/
The AI self-driving car could also be interacting with the occupants during the PRT ride time. It is generally assumed that the inside of an AI self-driving car is going to be outfitted with elaborate electronic entertainment options. You are going to be in your self-driving car quite a bit and have no need to be driving, and so the odds are that you’ll be doing some kind of game playing or maybe some kind of online educational learning, and so on. All of those aspects can still be underway while you are in the PRT.
In fact, in one sense, the human occupants of the true AI self-driving car might not know or even care that they are inside the PRT system.
Here’s why. Let’s pretend you’ve told your AI self-driving car to get you across town. You leave it up to the AI to decide how to best get there. It calculates that the roadways are jammed. It therefore drives to the nearest pod station of the PRT, gets into a shell, whisks along to the destination, drives out of the shell, and takes you the last mile to your destination. You were the entire time absorbed inside the self-driving car with that exciting class on the philosophy of Aristotle that you are taking during your morning commute, and had no idea that for part of the ride that you happened to end-up in the PRT.
One criticism of PRT’s is that it still does not solve the last mile problem. In other words, there are going to be only so many of the podcar stations around town. The stations aren’t going to be at every block and intersection. So, you’ll still need a means to get to wherever the access points are for using the PRT, doing so from your home or from your office or from the store, etc. The AI self-driving car is one means to get that last mile. You could use something else, such as a scooter to get there (and then use a PRT provided podcar), but the concept will be that you’d normally use some more robust personal space device like a car or a self-driving car.
Some people say that with PRT, we’d need less cars. We’d need less conventional cars and we’d need less AI self-driving cars. That could be true. That’s certainly the intent of the mass transit systems. Presumably, the same should be true of the PRT. Some don’t like the idea of the PRT allowing cars as the podcars, because they believe it merely fosters a car culture. They say let’s get rid of cars. We need to discourage cars, rather than encouraging them by finding other means to keep them around.
Others believe that the PRT might aid the expanded use of AI self-driving cars. The AI self-driving cars will be routinely going back-and-forth from people’s home or work or whatever, and going to the PRT, like bees going to a beehive. There are some that think it will foster more carpooling, since you and I might live next door to each other, but work in different parts of town, but we both get into our jointly owned AI self-driving car (or someone else’s ridesharing AI self-driving car), it uses the PRT to get close to my office and then drops me off, and then gets back into the PRT to do the same for you.
It’s hard to say whether and to what degree the PRT will discourage or encourage car use, whether conventional car or AI self-driving car.
Is there any chance at all that a PRT like this will even be built and be put into use?
Elon Musk believes so. He recently spoke here in Los Angeles and provided his latest vision of a PRT that he calls the Loop. You drive your car, hopefully in his case a Tesla, into a shell, doing so at ground level, and an elevator takes it underground. There is an elaborate tunnel system underground that acts as the network of the PRT. Once you reach your destination station, the elevator lifts the car back to ground level and you drive off. He’s drafted a map of 60 miles of tunnels with about 23 stops that his Boring company would create here in Southern California.
Critics decried his vision as impractical and unworkable. There will be long lines of cars at the pod stations, all waiting to get into a pod shell. Cars will be backed up for miles and it will be worse environmentally and disruptive to anyone working or living near a pod station. The places where the pod stations are located have been criticized as not enough of them and they are at places that seem randomly chosen rather than chosen for purposes of reducing travel time for those living here. Some said he ought to help improve the mass transit system here, rather than avoiding doing so by concocting something else entirely. And so on.
Maybe his vision needs tweaking. Or, maybe, like anything that’s a first try attempt, it needs more refinement. Or, maybe indeed, as some accuse, it’s a wacky idea. No one really knows. Other attempts at PRT’s have been met with resistance, some have met financial ruin, some have been unable to navigate the regulatory approvals, some had engineering design flaws. You name it.
That being said, I don’t think that whether or not there are PRT’s especially impacts the progress on AI self-driving cars. I figured you’d be relieved to know that. I do though consider the PRT to be another interesting “edge” problem for AI self-driving cars, meaning a problem not necessarily core to AI self-driving cars, but one that is nonetheless worthy of attention for AI self-driving cars. I’ll let you know when we get our AI self-driving car to be one of the first to drive on the SoCal PRT and will post pictures of it getting into and out of the pod shells, when or if that day arrives. Meanwhile, we’ll all need continue enjoying driving our PNRT’s, Personal Non-Rapid Transit conventional cars.
Copyright 2018 Dr. Lance Eliot
This content is originally posted on AI Trends.