I once had a communications expert from Toyota tell me: “You can tell how long someone has been in this business by how quickly they think self driving cars are coming.” – and as I’ve studied the subject for the last 2 years I’ve realized, he’s right! mostly. So lets just into the question:
just how long is it going to take for AVs to get here, and when will we all have one?!
well, short answer: by the time 2022 has rolled around you will probably have ridden in one, but its going to be a decade or two (think 2025 if you’re enthusiastic about these things and 2050 if you’re not).
This question has to be divided into two aspects: 1) when are self driving cars going to be commercially available and 2) when will I have one. Lets tackle (1) first:
When are self driving cars going to be commercially available?
Well, I went and tried to gather press clippings and market chatter and this is what I came up with:
So you see, it looks like somewhere around the turn of the decade we will be seeing the cars becoming commercially available for you to go and buy one! This chart is always in need of an update because things change rapidly in this industry, but it helps give a feeling of the time line. And a quick note about that red-line cross out there by the GM numbers: As I explain here, the first AVs are going to be rolled out commercially, through ridesourcing services. So as they come online, you, yes- YOU! will have the opportunity to hail a ride from a driverless vehicle. Some segment of the population will come to rely on ridesourcing from AVs to get around, but there are a few problems that might come along with that: such as the potential elimination of transit (a really bad idea) and massive roadway congestion.
There is a wildcard worthy of mention here: Tesla. As we learned here, Tesla wants to use a Camera/Strong A.I. solution, so maybe they come up with a system that works sooner than everyone else, and you can buy an autonomous Tesla in the next few years, but for the vast majority of the population, Tesla is probably not an option.
So in no way minimizing the massive impact ridesourcing services are going to have on the AV rollout (you may never own another car again). Lets still take a look at more traditional commercial, everyone buys their own AV model.
When will everyone have an AV?
Lets start with a base case: the average vehicle on the road has a lifespan of 11 years. So lets say we flip on the “autonomous” switch tomorrow and every now car tomorrow. That means that there will be at least 11 years before we got rid of the vast majority of human driven vehicles on the road way. That puts us around 2028. Ok, not too bad, but there is a problem with that: These cars aren’t going to be here until the turn of the century: Ok, so add 3 and we are at 2031. Still not to bad! Well, there’s another problem: Currently Half of the Population is Not Interested in Autonomous Vehicles. Here look at this study TTI did, where they surveyed 556 Austin residents:
50/50! wow. So, what does this do to our prediction as to when we will all have our own AVs – well, it makes it muddier, I went through a culled a bunch of different predictions from various sources around the internet and included them here below:
Source 1: Victoria Transport Policy Institute
By comparing the adoption of past auto technology (like electric vehicles, airbags and electronic steering control) you can estimate that autonomy will follow similar adoption patterns. From that VTPI came up with 60% penetration by 2055:
Source 2: Consultants!
All the big consulting firms are working on coming up with predictions for their clients. I looked at a bunch but, for brevity want to highlight two here. First McKinsey has a prediction that looks like this:
So they do a great job recognizing that this could happen fast or it could happen really slowly. Of course “fast” here means 50% around 2033 – so that’s not a crazy distance from our back of the envelope calculations above.
In this category, I also wanted to recognize Morgan Stanley:
Do you remember how I said you can tell how new someone is to the industry based on how fast they think autonomy will get here? Well, this prediction, made a few years ago, illustrates that concept! Morgan Stanley predicts an autonomous society in 2 decades (2040s), and at the same time, they anticipate such cultural upheaval, in a good way, that we also arrive at this utopian society around the same time. Let me just say: I think they are wrong (too fast), I hope they are right, and I love their optimism!
Source 3: Academia
There are many academics working on this project but we’ve got a few really good ones at UT-CTR. One in particular to highlight here is Dr. Kara Kockelman. She has put a lot of effort into studies that really help direct the philosophy on the adoption trends and wider effects of AVs. She, with Bansal, conducted a bunch of surveys and came up with a range of potential adoption curves, graphed here:
So we see, its going to be a few decades, by her work before we start to see this widespread penetration.
So there you have it! To recap:
- AVs will be here around the turn of the decade
- Ridesourcing services will have them first and could play a hugely disruptive role in how fast we all use AVs
- Commercial availability (for personal purchase) will take decades to filter throughout the rest of the population
thanks for reading!