By Lance Eliot, the AI Trends Insider
One of the most challenging aspects of ridesharing involves the empty roaming time.
In short, human drivers are continually underway in hopes of being in the right place at the right time in order to be the car most likely chosen for a ride request. Since the ride-sharing firms oftentimes aren’t on the hook for the effort and cost associated with cars that are roaming, the firms don’t especially care as much about the idle time of those drivers. The primary focus is on making sure that the shortest wait time can be achieved for the riders making requests, or those paying passengers might switch over to a competing ride-sharing network.
Sure, there is a concern too about those drivers due to the aspect that a driver that seems to endlessly roam will not be desirous of listing with the ride-sharing network that leaves them high-and-dry. In that sense, it behooves both the drivers and the ride-sharing operation to try and minimize the empty roaming time and seek to maximize the passenger riding time.
This is a complex problem overall.
There are some number of unknown requests that will arise, occurring from unknown locations, for which the aim is to have a driver and car that is relatively nearby when the request comes in. There is (generally) an unknown number of cars and drivers, along with an unknown dispersal of where those cars and drivers are. A driver might be tempted to simply park their car and wait for a request, but it seems more proactive to keep underway and attempt to be in the midst of the action.
There is a natural kind of tension between the ridesharing firm making suggestions to the drivers as to where they ought to roam versus the drivers wanting to remain independent and use their own wits to figure out where to be. Some drivers seem to be either astute or lucky, happening to keep their plate full and avoid being in the wrong places at the wrong times. Others somehow are driving in a desert and rarely seem to get the pick of the rides.
It is a gigantic logistics issue and one that attracts top-notch data scientists and promises handsome rewards for algorithmically trying to find the best possible approaches.
Let it be said that one of the thorniest and problematic aspects that can stymie ride going services is the percentage of time that the passenger-hauling car is empty of passengers. Presumably, no money is being made at those times. Worse too, there is cost during those times, such as the cost of the fuel being consumed, the cost associated with the wear-and-tear of the vehicle, and obviously the cost associated with the driver labor. If the ridesharing firm is not directly paying the driver, the cost then is a lost opportunity cost for the driver rather than a cost to the networking firm per se.
Generally, in recap, a car being driven by, say, an Uber or Lyft driver that is bereft of any passengers is not a good thing.
You can pretty much assume that when there isn’t a passenger there also isn’t any income being derived by the ridesharing effort. One quasi-exception consists of instances in which the driver is delivering something, perhaps a pizza or a burger or a package, and as such, there is presumably income covering the delivery action.
For purposes of simplicity, let’s assume that delivering something is equivalent to having a passenger.
Thus, when I refer to an empty vehicle, it means that the car doesn’t have a passenger and nor is it in the act of delivery. Also, for clarity’s sake, a passenger could be a human being, though the rider could also be a favored dog or cat that someone is paying for a breezy ride to the vet or grandma’s house.
Another factor to keep in mind about a so-called “empty vehicle” is that there is a human driver sitting in the driver’s seat.
Sometimes it is confusing to refer to an empty car since it could imply that nobody is in the vehicle at all. In the context of ridesharing, an empty vehicle traditionally means that there is a driver and no one else in the vehicle.
I’ll further point out that there might be a passenger in a car, but if they aren’t paying anything to be there, they essentially count as an empty vehicle. Maybe a ridesharing driver decides to give a friend a free ride to the store, or perhaps the other person is a fellow driver that might be in training. If there’s no dough derived from the passengers, it’s considered an empty vehicle.
The formal wonky moniker is often referred to as the utilization component of ridesharing or e-hailing service, typically calculated as the percentage of time that a driver spends with a passenger. We can quibble a bit about whether emptiness and utilization are the same or not, but for sake of discussion let’s assume that they indeed are one and the same.
Industry Average Of Empty Car Time
Can you guess what the industry-wide average of empty vehicle time is?
When I ask this question at conferences, attendees toss out all kinds of wild numbers.
Some think it must be rare that a ridesharing car is empty and so guess what it is maybe 1% or 2% of the time. Others guess that it might be relatively common to have an empty ridesharing car, and they guess that it might be the seemingly astronomical amount of say 20% or one-fifth of the time.
According to industry reported stats, the average time that a ridesharing vehicle is “empty” comes to about 41% of the time.
I hope that wasn’t overly shocking for you (I should have said trigger alert).
Though the stats don’t also say how much of that empty car time involves the car being in-motion, most would assert that the ridesharing cars are nearly always in motion when available to provide a ride.
Generally, when a ridesharing car is parked, it implies that the car is not available for a ride. There are exceptions, of course, but the gist is that the 41% implies that a significant chunk of the time that a ridesharing car is underway, it is empty.
The emptiness factor produces a lot of ugly and undesirable consequences.
A ridesharing car that 41% of the time is not earning money means that the cost of the driver and the cost of the car are having to be covered by the 59% of the time that a passenger exists. Presumably, if you could decrease the emptiness, you could produce more income, and in theory pay, drivers more and more readily cover the costs of the car, along with gaining a profit.
Another aspect that gets people mad about the empty car phenomenon is that the vehicle is adding to traffic snarls, presumably exacerbated by the wasted 41% of the time there isn’t a passenger in the car. If you stand on the street corner of a busy city, you can likely see a parade of ridesharing cars after ridesharing cars that are cruising the streets, absent a passenger, and trolling or waiting for a passenger to hail the vehicle.
Anyone driving their car for their own purposes often gets steamed to see themselves surrounded by empty ridesharing cars. What might have taken ten minutes to drive from your office to a restaurant gets extended multifold by those darned empty ridesharing cars that are choking up traffic?
Cities have qualms too about the pollution that those empty cars are producing. It’s one thing to be driving a passenger and producing pollution but driving an empty vehicle and gushing out pollutants seems especially egregious and foul.
An often-unstated downside of roaming empty cars is that presumably the risk of getting into a car crash or other incident is increased. This risk is not due to the car being empty, but instead due to the length of time on the roadway and the distance traveled. The more any car is on the roadway, the chances of getting into an accident continue to be present.
Some question whether the safety compromise of having roaming ridesharing cars that are empty is warranted to the public at large, though few studies exist to predict how much of the 41% of that overall driving time is leading to car accidents (other than using generic driving stats as an approximation).
Kerfuffle In NYC
A bit of a kerfuffle rose last year in New York City (NYC) on the emptiness conundrum. The NYC Taxi and Limousine Commission (TLC) passed a regulation that required ridesharing firms in NYC to achieve a lowered rate of emptiness, striving for a 31% (or less) goal.
On the surface, aiming for a lessened emptiness factor is a good thing, though drivers are typically worried that it means they might be forced to cut back on their driving efforts and therefore earn less money, and the ridesharing services are concerned about backlash from all quarters.
For example, suppose that the time for you to wait once you’ve requested a ridesharing ride goes up, which could happen if the number of cruising empty cars is reduced to knock down the emptiness time. As a customer, you aren’t likely to be pleased that the emptiness factor went down and yet the delay for you to get a ride went up. All you’ll care about is the wait time. Most ride-seeking people are bound to say that they don’t care about emptiness and only care about getting a prompt ride.
You can argue that maybe their journey in the ridesharing car might be faster, and they might end-up shell out less money for the ride once the emptiness rates drop to 31%, though right now that’s a theoretical proposition and the actual impact could come out quite differently.
Here’s an interesting point to consider: Will the advent of self-driving cars eliminate the emptiness factor, or will it be the same or perhaps even worse than with today’s conventional cars?
Let’s unpack the matter.
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/
Self-Driving Cars And Ridesharing
True self-driving cars are ones where the AI drives the car entirely on its own and there isn’t any human assistance during the driving task.
These driverless cars 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 considered semi-autonomous, and typically contain a variety of automated add-ons 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 point out).
Since the semi-autonomous cars require a human driver, akin to any of today’s conventional cars, I’m not going to consider the emptiness factor a Level 2 or Level 3 car (it would be the same as the emptiness of conventional cars). Instead, let’s focus on the emptiness or utilization aspects involving true self-driving cars, ones at Level 4, and Level 5.
Well, the first big difference is that when a true self-driving car is considered empty, it really is empty.
Recall that with conventional cars we were willing to say that a car was empty when there weren’t any passengers, even though there was a human driver present. Now, due to the AI driving system, there isn’t a human driver and presumably not a provision allowing for a human driver (this is a controversial point, for which some believe that human driving ought to be still allowed in true self-driving cars).
From the perspective of a ridesharing or ride-hailing firm, removing the human driver is somewhat like a godsend. No more dealing with those darned human drivers that are cantankerous and complaining about the money they are making or being denied. Human drivers oftentimes pick and choose when they want to work and dare to take off days whenever they feel like it.
Overall, the headaches of human drivers are erased.
Furthermore, the AI driving system can drive whenever and wherever the owner of the self-driving car dictates. There are no debates about working late at night or having to drive to Timbuktu for fares. Instead, the owner deploys the self-driving car to any place and at any time.
Many are tempted to also say that the cost of the driver is removed. Yes, certainly the hourly fee paid to human drivers or commissions is no longer pertinent, but you need to impute the cost of the AI driving system as a kind of surrogate for the cost of a human driver. The AI driving system consists of the plethora of sensors on the self-driving car, plus the computer processors, plus the AI software, and so on. The hardware is prone to wear-and-tear. The software will need updates.
All in all, there is still a form of “cost” associated with the driving act. Nobody knows what the cost of these AI driving systems will be.
It will be interesting to see what happens when true self-driving cars start to compete head-to-head with human-driven ridesharing and ride-hailing cars. We will begin to know how the costs compare and whether the AI driving systems will indeed drive down the costs as most assume or hope will happen.
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: http://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/
Self-Driving Cars And Emptiness
Returning to our focus on the topic of emptiness, there is nothing magical about self-driving cars that obviates the emptiness factor.
You can have self-driving cars roaming around and cruising the streets, doing so without passengers, and in the same manner that human-driven “empty” ridesharing cars do so.
If that’s the case, what do you predict the emptiness percentage will be? It is tempting to say that it might be about the same as today’s 41%. Maybe, maybe not. Some worry that it might be a lot higher.
Here’s why that could happen.
Everyone that owns a true self-driving car is going to want to wring every dollar of ridesharing revenue they can out of the driverless vehicle. Might as well put the self-driving car into service nearly 24×7, minus the time needed to refuel or recharge the car, and minus the time needed for maintenance or repairs. Imagine then a myriad, nay a swarm, consisting of hundreds or maybe thousands of self-driving cars roaming around, all waiting for that moment when they will be called into action to carry a passenger.
Furthermore, pundits keep saying that we can get rid of parking lots in downtown areas since the advent of self-driving cars will make parking lots no longer needed. You would be stupid or outright foolhardy to park your self-driving car when it could instead be picking up and transporting passengers.
At least that’s the conventional wisdom about our future.
Some pundits are quick to gush about the fact that downtown areas will be able to reclaim parking lots for more important uses such as additional housing, or businesses, or retail space, or green space and trees. If you insist that parking ultimately is needed for those self-driving cars, the answer glibly stated is that parking in areas outside of downtown could be set up, doing so in places that the land is much cheaper and no humans live or want to live nearby.
Consider once again the emptiness factor.
You’ve got zillions of self-driving cars roaming around the streets of downtown, and for some percentage of the time, those self-driving cars are empty. Also, those self-driving cars are going to be driving empty to an outside area that has a parking lot. According to those pundits, no humans are going to be getting a lift out to those parking lots because the self-driving car parking space will be on the outskirts where no humans are desirous of the land.
When you start to add-up the emptiness time, it could be that self-driving cars will break the sound barrier of emptiness and reach astoundingly high percentages.
For those that own self-driving cars, the emptiness is going to hurt and could make the use of the self-driving car for ridesharing unpalatable.
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: http://aitrends.com/selfdrivingcars/self-driving-car-lawsuits-bonanza-ahead/
Tackling The Emptiness
Ponder some of the adverse consequences already mentioned about emptiness and recast them into a world that includes true self-driving cars.
Here are some key points:
- The traffic snarls caused by self-driving cars vying for fares could be worse than what we already experience, partially due to the urgent need by the owners to seek out revenue to cover their expensive self-driving car purchases, and ongoing maintenance, and therefore sending their vehicles into an infinite search for paying riders.
- Human drivers that are also trying to drive, whether for ridesharing or personal purposes, are certainly going to get irked at seeing all those headless self-driving cars jamming up the streets. Might human drivers’ rebel against their AI brethren?
- Passengers seeking a ride might have elongated wait times, especially if the predicted induced demand materializes (induced demand means that people today that don’t use a car might be induced to do so, once we have self-driving cars).
- Journeys inside a self-driving car might be pleasant due to being able to relax and watch TV, but if the trip across town takes twice as long as it used to take, will riders be happy with the result?
- On a safety basis, once again the longer a car is in-motion and consuming miles, the risks of getting into a car accident continue unabated. Those that live in a Utopian world and insist that self-driving cars won’t get into car accidents are making a false assumption that only self-driving cars will be on our roadways. We today have 250 million conventional cars in the United States alone and they aren’t going away anytime soon, so stop pretending that we’ll only have self-driving cars and realistically expect that we’re going to have a mixture of human-driven cars and self-driving cars for a long time ahead.
- Pollution might be the one factor that does get impacted in a good way, namely that self-driving cars are likely to be Electrical Vehicles (EV), therefore the amount of pollution being pumped out of the cars will be less than today’s gasoline-powered engines. The flip side to that coin is that there is a pollution footprint produced in the generation of electricity that is needed to recharge EVs.
Some say that we should restrict downtown areas to only self-driving cars and refuse to let human-driven cars into those areas.
In doing so, perhaps the emptiness factor might come down.
The logic being that if you have less of a supply of ridesharing vehicles in the restricted area, the demand is going to fill-up the self-driving cars more so. Of course, it might have the added effect of increasing wait times to get a ridesharing lift.
Will human-drivers be willing to get shoved down on the totem pole and sit on the outside while self-driving cars get the vaunted and exclusive access to making money in the steeped ridesharing locales such as a downtown area?
You can already see the ridesharing human drivers protesting that the small guy is getting the shaft, while the big businesses that own fleets of self-driving cars are getting outrageously wealthy.
I’m not trying to paint a doomsday picture concerning the role of self-driving cars. My overall point is that some believe that the advent of self-driving cars will be a boon for mobility and make car rides readily accessible to all. Though I am hopeful and optimistic about such an outcome, we need to realize that the real world is going to intrude on how things will play out.
Empty cars, whether for human drivers or self-driving cars, do not provide a free lunch. If a ridesharing car is not carrying a passenger or a package, presumably there is no one paying to have the resource meandering along on our streets. Self-driving cars are a resource and have an associated cost.
Trying to figure out how to best manage that resource will be an issue for fleet owners and individual owners, along with being an issue for cities and for people that simply want to get a ride.
It’s going to be a challenge, and hopefully, one that we can avert the woe.
Copyright 2020 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/]