Pet Mode for AI Autonomous Cars

Transporting of pets in the car can be a dicey proposition.

By Lance Eliot, the AI Trends Insider

What is the most popular type of pet in the United States? Kind of a trick question, I suppose, since you undoubtedly first thought of dogs or cats. The answer is that freshwater fish are the most popular pet, consisting of an estimated 142 million of them.

I’ll give you another try: Are there more dogs or cats as pets in the United States? 

Turns out there are about 88 million pet cats and 75 million pet dogs in the US, so cats come out to be the victor in terms of popularity by count. Dog owners would likely argue though that in spite of there being more cats, maybe we should count popularity by some other factor and dogs might therefore be considered the top dog, so to speak.

Don’t Miss It: Zombie-Car Taxes Are Arising For AI Autonomous Cars

Putting aside the debate about which kind of pet is the most popular, it is somewhat surprising to realize how big the pet market is. 

An estimated 68% of United States households have a pet. The spending on pets in the United States is an estimated $70 billion or more. That’s a lot of money. That’s a lot of households. That’s a lot of people that either own a pet or maybe enjoy being with someone else’s pet. If tomorrow somehow all pets suddenly disappeared, think about how it would impact people’s lives and how much we’ve come to rely upon having at the ready our pets.

I used to have a dog. I used to have a cat. I mention both of them since some of you might think I’m more of a dog-person or a cat-person – I’m an equal opportunity pet owner (I’ve also had freshwater fish, birds, reptiles, etc.). For my dog, I would occasionally take him down to the beach for a romp on the sand. He loved to run around and chase the birds and chase the waves as they crashed on the beach. It was quite a workout for him and he’d be tuckered out by the time we got back home.

 Getting him to the beach was a bit of a hassle. 

Animals Inside A Car Can Be A Daunting Matter

At the house he roamed free. When I took him in the car, I’d put on his leash (he usually assumed he was going for a walk around the block), and then I’d put him into the backseat of the car. Actually, the moment I opened the door of the car, he knew what was going to happen next and with grand delight he’d leap into the car. I’d use the leash to loosely tie him down so that he couldn’t wander throughout the car. As with most dogs, he enjoyed putting his nose outside the window to smell the cacophony of odors as we drove to the beach and so I always cracked open the window for him.

 Excitability would sometimes get the better of him while in the car. 

This meant that there might be biological emissions during his time in the car, including excrement and urine, which obviously is not what one usually hopes to have in their car. Anticipating these moments, I put towels in the car and a blanket upon which I hoped he would generally stay, plus I tried to train him to “hold it” until we got out of the car at our destination. The return trip from the beach was similar to the trip to the beach, except that he usually was tired and would lay down in the car. This had its own disadvantages because he often had sand on him and other muck that he might have encountered during his beach romp.

The joy of playing with him at the beach, and seeing his joy of being at the beach, made the whole gauntlet of steps to do the drive worthwhile. For my cat, trips in the car were almost always solely to take the cat to the vet. Unfortunately, the cat figured this out. As a result, the moment that I started to make motions that I was going to take the cat to the vet, the cat would hide or play hard to get. I tried a few times to let the cat be loose in the car, having a leash similar to what I had done with the dog. The cat though hated being in the car and would scratch and hiss, so the best means of transport ended up involving the use of a pet carrier for the cat (it wasn’t so much the car that the cat hated, as it was the realization that it was time to see the vet).

Whenever you have an animal inside a car, it can be a dicey proposition.

As the driver of the car, you certainly don’t want the animal to interfere with your driving. 

A cat that’s allowed to wander anywhere within the confines of the car could suddenly jump in your lap and you’d be so startled that you’d maybe steer the car off a cliff. A dog that can move around could get angry at a dog on the street and start barking, maybe distracting you, doing so just as you are making a right turn and perhaps you inadvertently hit a nearby pedestrian.

If you are transporting freshwater fish, I suppose it’s less likely of something going amiss in the car, though if you have them in a simple fish bowl, and if you happen to hit the brakes while driving, they might go flying throughout your car, and in so doing disturb you that you become distracted from the driving task. 

Anything can happen.

A rule-of-thumb would seem to be that for any animals inside a car, it’s best to control them in a manner that they cannot disturb the driving of the car.

I had a friend that would carry his dog in his arms and ride as a front-seat passenger in someone else’s car. One day, the dog freaked out when another car honked its horn, which caused my friend to reflexively try to re-grab his dog, which caused my friend to flail around in the passenger seat, which then he unintentionally hit the driver, and the driver of the car then rammed into a car ahead of them. Quite a story to tell the police or the insurance company.

The story is helpful because it highlights that there is a range of “control” that one might have with the animal. Some people tie down their pet while it’s in the car and try to immobilize it from impacting the driver. Some think that verbal commands alone will keep the animal from getting out of hand. My friend thought he could just hold his pet in his arms. My having put the cat into a pet carrier pretty much prevented the cat from disturbing my driving, though the cat meowing and hissing did admittedly perhaps distract me somewhat from being fully attentive to the driving task.

 Impacts Of Having An Animal Inside A Car

Here’s some possibilities that we want to presumably avoid when transporting an animal inside a car:

  •             Animal endangers the driver
  •             Animal endangers other occupants
  •             Animal endangers the car
  •             Animal endangers itself
  •             Animal endangers others outside the car

 I saw a dog leap out a window of a car one day and chase after a person that the dog apparently disliked, thus, it’s conceivable that an animal could endangers others outside of the car. At supermarkets, you sometimes see people that have parked their cars and left an angry dog in it, which when people walk past the car, the dog tries to take a bite out of them. Not good for anyone, humans nor pet.

 There have been cases of animals inside a car that wreaked havoc on the interior of the car. 

I even heard one time that a person left their engine running, got out of the car to do some quick task, and the dog somehow bumped into the transmission knob and the car went into drive.

Luckily, miraculously, I don’t believe anyone was hurt in that instance. I did hear though that the dog went to the Department of Motor Vehicles (DMV) in hopes of getting a driver’s license. 

Well, maybe not.

I know that some people get upset when I say that you should control your pet while it’s in your car. 

They argue with me that if the animal is domesticated, there should not be any concern about controlling the animal. Only if the animal is a wild animal do they think that there’s any need to be overtly thinking about controlling the animal. I don’t want pet owners to get upset with me, but I’ll just point out that even the most domesticated animal still has animal instincts and reactions.

In my view, you can’t be too careful, especially when it comes to putting an animal into the rather confined space of a car and once the car gets into motion you are opening up a can of worms. It can be an explosive and dangerous combination.

Pets can do any of these things:

  •         Become scared
  •         Get startled
  •         Become upset
  •         Get angry
  •         Try to run or scamper
  •         Bite
  •         Become confused
  •         Scratch
  •         Etc.

 If anything, I’m often concerned not just for the human driver or the human occupants, but also for the animal itself. When someone thinks they are doing a good thing by letting their dog roam freely in a moving car, they aren’t thinking about the harm that can come to the dog. Suppose the driver suddenly hits the brakes? That dog is going to go flying in the car and possibly get injured or killed by hitting something in the car or maybe even getting thrown outside of the car. Most states have various laws and regulations about restraining your pet while it is in your moving car, for its safety and your safety.

 Autonomous Cars And Pets Going For A Ride

What does this have to do with AI self-driving cars?

At the Cybernetic AI Self-Driving Car Institute, we are developing AI software for self-driving driverless autonomous cars. 

This includes encompassing various “edge” problems such as the transporting of pets (animals) while in an AI self-driving car.

Let’s consider some of the aspects involved in this edge problem. I call it an edge problem because it is considered by most of the automakers and tech firms as something outside the core of what an AI self-driving car is supposed to do. They are focused on getting the AI to drive the car. The aspects of dealing with any pets inside a car is considered secondary and much lower on the list of crucial things to get done.

 For my article about edge problems in AI self-driving cars, see:

 For my framework about AI self-driving cars, see:

 Let’s also define what is meant by an AI self-driving car. There are various levels of self-driving cars. At the topmost level, Level 5, it’s a self-driving car for which the AI can fully drive the car. This means that there is no human driver required in the self-driving car at a Level 5. For the levels less than a Level 5, there is a need to have a human driver present. The human driver and the AI are considered co-sharing the driving task, though the human is also considered ultimately responsible for the actions of the car. This notion of co-sharing the driving task is something I’ve mentioned many times is raft with various drawbacks.

 For my article about the co-sharing of the driving task, see:

 For the levels of self-driving cars, see my article:

 In the case of an AI self-driving car that is less-than a Level 5, there needs to be a human driver at the ready to drive the car. This implies that if you do have an animal in the car, it’s similar to the situation today of having an animal in a conventional car. Anything that the animal does that disturbs the human driver can have adverse and dire consequences.

 I realize you might be thinking that if the AI is co-sharing the driving task, why doesn’t it just wrench control from the human driver if the human driver suddenly becomes unable to do the driving task.

 This has several problems.

 One is that how will the AI realize that it is best to take control from the human driver? Even if the AI is detecting whether the human driver has their hands on the wheel or maybe via a camera whether the human is looking forward at the road, trying to judge when it is appropriate to take over control is not very transparent. Imagine too if the AI does take over control and it turns out that the human was still in control, but that the AI now maybe is going to take an untoward action since it doesn’t know what the human driver was intending to do.

 You could even get into a tug of war between the AI and the human, trying to take control from each other. I suppose you might contend that if the human tries to take control back, the AI should relent. But, suppose the human doesn’t know now what the AI was intending to do, and so the human puts the car into an untoward posture. Or, maybe the human takes back control, but then let’s say the dog in the car jumps on the human a second time and it is necessary for the AI to once again take control. Meanwhile, maybe you’ve used a rule that if the AI takes control, and the human takes it back, leave the control with the human, avoiding a tug of war. That wouldn’t work out in all cases.

 You might say that if the AI is indeed watching what’s going on in the self-driving car, maybe it should be sharp enough that it can figure out that say a dog is loose or cat is loose. Yes, you could potentially have the machine learning recognition that would be programmed for this, but it is much subtler than you might think. Generally, it would require some kind of common sense reasoning to try and decide whether the dogs or cats are doing something relatively safe or unsafe.

 For my article about common sense reasoning in AI self-driving cars, see:

 Another perspective is that you might have the human driver tell the AI to take over control. This would be similar to when you initiate say Alexa and so you say a code word, “Alexa” and then you provide some kind of command. It could be that the AI self-driving car has a code word, let’s use “Lance,” and then after you say that word you can tell it to take over control of the driving task. I realize that some of you will say that suppose a child in the car suddenly yells out the code word and then all of a sudden the AI construes whatever is said next as a command (“go off a cliff”). Some counter-argue that the AI could be using a voice fingerprint identify capability such that it would only recognize and acknowledge the human driver’s voice at the time of the driving task.

 In recap, here’s some aspects involved:

  •         AI tries to ascertain if an animal is amuck
  •         AI potentially takes over driving task if human driver seems unable
  •         Human driver can signal to the AI to take over the driving task
  •         Human driver can signal to take back the driving task or refuse to give it up
  •         Miscommunication in the co-sharing could have dire consequences
  •         Misunderstanding in the co-sharing could have dire consequences

 Pet Mode For An Autonomous Car

One approach too involves having the AI be a kind of an alert monitor about the animal in the self-driving car. 

In essence, the AI could be placed into “pet mode” and be ready for the particular dynamics of having a pet inside the self-driving car. This pet mode could be initiated overtly by the human occupant telling the AI to go into pet mode, or it could be figured out by the AI via machine learning recognition of the “objects” inside of the self-driving car (and possibly with a confirmation to the human driver that indeed there is an animal on-board).

Part of the “pet mode” could be that the AI would be on the watch on behalf of the human driver about things that the pet is doing. It would be akin to having a passenger in the car that can tell you that the dog just chewed the backseat armrest, or the cat is curled up in a ball on the floor. With a human passenger, they would be able to take physical action when needed and help restrain the animal, but there’s not much the AI of the self-driving car can do in that regard. Instead, the AI would be devoted to warning the human driver about what the animal is doing, and being another pair of eyes, so to speak, while the human is presumably watching the road and being at the ready for the driving task.

You could potentially have the AI try to talk to the animal. Perhaps your pet dog has been trained to listen to the AI of your self-driving car. The AI then could potentially tell the dog to sit down or get into the backseat. I realize this seems kind of wild as an approach, but as you’ll see in a moment, maybe it’s not as crazy as it seems.

Let’s now consider the use of “pet mode” for a Level 5 self-driving car. In the case of the Level 5, the AI is doing all the driving. There is no human driver. This is handy because it implies the animal is unable to distract the driver. Whatever the animal does, the AI is still going to be able to drive the car.

The only way the animal presumably can disrupt the driving would be if it is able to damage something inside the self-driving car that could hamper the AI or the car itself. Suppose that there is wiring just under the dashboard and somehow the dog gets to it and chews through those cables. If they are involved in the electronics of the car and any kind of driving related task, it could be disruptive to the AI and the car integrity.

This then brings up facets about the interior design of the self-driving car. It is generally envisioned that since there is no longer a need for a human driver in a Level 5 self-driving car, we can redesign the interior compartment of the self-driving car. No need to have a seat facing forward in the same place that today’s driver seats reside. Instead, the compartment can be perhaps seats that swivel and face each other. Or, maybe seats that can recline fully so you can sleep in your self-driving car when you want to do so. With the potential of true AI self-driving cars being used non-stop 24×7, it is presumed that people will likely sleep while on their way to work or on trips to visit in-laws, etc.

For my article about non-stop AI self-driving car use, see:

When redesigning the interior of cars to be suitable as a self-driving car, one additional consideration will be the nature of the occupants and what they might do inside the self-driving car. If you are going to put your children into a true AI self-driving car in the morning so that they will be driven to school, doing so without any adult supervision inside the self-driving car, you want to know that the children hopefully cannot harm themselves by poking around within the interior of the car. Today’s cars leave all sorts of metal joints and prods fully exposed inside the compartment, which a small child without supervision could easily harm themselves on. The same could be said about pets that are unsupervised.

Indeed, when you consider that 68% of U.S. households have pets, and once there are true AI self-driving cars prevalent, just imagine how many of these households will opt to send their pet by itself in the family AI self-driving car to go visit the vet. Or go visit grandma. Or go to a pet playground where there is someone to supervise your pet, and then they put your pet back into your AI self-driving car, which comes to work to pick you up at the end of the day, and your joyful dog is right there to greet you. No need to wait until you get home to get some hugs and kisses from your beloved pet.

The redesign of the interior of a car should take into account the notion that today’s designs are inherently dangerous and assume that whomever is in the car will be somewhat supervised. True AI self-driving cars won’t necessarily have an adult in the car to undertake supervision of children and nor pets. Overall, this implies that the interior has to be made safety proof with regard to whomever is inside the self-driving car. There shouldn’t be any easy way to cut yourself. There shouldn’t be any easy way to undermine the capabilities of the self-driving car by hitting something or chewing on something.

This does not mean that the interior needs to be a steel tank or barren fortress. The automaker can provide various covers and shields to allow for a relatively impervious interior. I’d even assert that if the automakers don’t do so, a thriving third-party market will likely develop to outfit your interior compartment so that it is safer for the transport of children, pets, etc.

I had mentioned earlier that the AI might be made to talk to the pet. Suppose your pet will be in the true AI self-driving car for an hour or two, perhaps taking a lengthy journey. That’s a long time for your pet to be alone. The AI could be talking to the pet, maybe playing music, or otherwise try to comfort the animal. Given that there are likely cameras pointing inward, you can do a Skype like chat with your pet, and the inside of the AI self-driving car is likely to have screens, usually used to show movies or do your digital work. Your pet could see you, you could see your pet, and try to comfort your pet during part of its journey in the self-driving car.

I had mentioned earlier that the seats in a true AI self-driving car might swivel and there might be other variations in terms of configuring the internal compartment. For those that have pets, there might be ways to re-configure the inside compartment to allow for taking out the seats and allowing the pet to wander around inside the car.

Perhaps there might be a special leash or restraint system that keeps the animal relatively safe, but also allows for open movement while in the car, most of the time (the restraint system might opt to more strongly restrain the animal if getting into rough traffic, or maybe if the pet is getting out-of-hand). Anyway, you can just see the ads now, the XYZ auto maker comes out with a pet friendly AI self-driving car and tries to lure buyers that have pets – could be a sizable segment of the market.

For my look at how marketing of AI self-driving cars is likely to occur, see:

It is anticipated that the advent of true AI self-driving car is going to be a tremendous boon to the ridesharing industry. People will buy a self-driving car, realize that it can be used 24×7, and opt to rent it out while they are at work or asleep. Other people will buy a self-driving car solely as a ridesharing revenue maker and not even use it for personal purposes. We’re heading toward a ridesharing economy, or ridesharing-as-a-service world.

That being the case, what about pets? People are going to want to have their pets go lots of places that right now involves too much of a hassle for them to drive their pets to, and yet with a true AI self-driving car it could be a breeze. Some ridesharing services might tout that they have pet-devoted AI self-driving cars, ready for the transporting of your favorite dog, cat, or fish. I’ve already predicted that with the advent of AI self-driving cars we are going to see all sorts of induced demand. This is demand for using a car that otherwise today is suppressed or that people don’t even think about currently.

For my article about induced demand and AI self-driving cars, see:


I began this discussion by pointing out that we today spend $70 billion on our pets, doing so for pet food, pet toys, pet care, and the like. It seems logical and inevitable that true AI self-driving cars are going to ultimately intersect with our desire to have pets.

Right now, if I told you that someday your pet will be driven around in an otherwise empty car, you’d think I was loco or that it would have to be some crazy rich person that is spending way too much money on their pet. In the future, the prevalence of AI self-driving cars will open the avenue for considering ridesharing involving our pets. Easily, readily, at a low cost. It’s going to happen.

The idea of a pet mode for an AI self-driving car is not particularly far fetched if you have a long-term view, at least that’s what we say.

Well, I suppose it could be that me and my team just love our pets so much that we insist on somehow getting them involved in AI self-driving cars. 

I wonder if I could train a dog to do AI coding? 

Or, would a cat do a better job at it? 

It’s hard to say.

Copyright 2019 Dr. Lance Eliot 

This content is originally posted on AI Trends.

This UrIoTNews article is syndicated fromAITrends