In this Q/A session with Aaron Nathan, founder and CEO of portfolio company Point One Navigation, Aaron shares his perspectives and outlook on the future of autonomous vehicles. Point One Navigations is the most reliable GPS precision service for autonomous vehicles and unmanned drones.
Why should we be excited about autonomous vehicles?
The biggest thing about autonomous vehicles will be the fact they they will free up so much time for people to do other things — whether it is to relax or to learn things. To have a zero mile commute effectively — the amount of time you'll be able to get back is incredible. That's probably the first thing.
The second thing is that the number of lives that you'll be saving — in the US that number is around 30,000 — that's bigger than any disaster that I know of. I think that's something that's often overlooked because we all like to think that we're the best drivers in the world, but realistically, everyone is a bad driver at some point because we're distracted or sleepy or in any kind of non perfect state. A computer doesn't suffer from those things — computers don't get tired; they don't get confused; they don't drink and don't text. In the long term, there's no question in my mind that we can build a safer car than any human will ever be capable of driving.
What does the future of autonomous vehicle look like to you?
Well, I think the near future is going to be autonomous vehicles in specialized applications. I think you'll see some autonomous vehicles doing repeated routes in cities. I think you'll see in a similar timeframe autonomous or semi—autonomous long haul trucks carrying goods. In some areas, autonomous vehicles will used for rural transport in places that don’t necessarily have tracks for a train. It will help people from certain communities get access to things in different cities — whether that be food, healthcare, all kinds of different services.
In the longterm, I think there's definitely a future where the autonomous vehicle can reshape how we rethink cities and suburbs and that's very exciting. I also think that's 50 years out before we start to see those effects. Just in the same way that when we look at streets, how they were designed in Europe and how they were designed in the US, that's a function of technology and that's going to happen again when you see autonomous vehicles. Streets will be designed differently and I think more efficiently. We'll have more green space and less asphalt. And that's also exciting.
What are the biggest hurdles to autonomous vehicles becoming mainstream?
The biggest hurdle is a precision localization system that actually works.
Realistically, there's several problems. The hardest thing is dealing with uncertainty in a way that makes sense and that happens in all levels of autonomous vehicle development. So whether we're talking about knowing the position of a car in a lane or whether you're looking to understand why a vehicle is heading towards you — is it because you're on a divided highway? because the vehicle is a little over a lane? or is it because someone is actually doing something adversarial — and then you should take evasive action. That's a very very difficult thing to determine even for a human. A lot of times, we assume that people are going to do rational things. And so, encoding that kind of assumption, that kind of intuition into a robot is something that is pretty challenging. A lot of people are making great progress towards it, but we're not quite there yet.
Apart from transportation, how do you see autonomous vehicles disrupting other industries?
Transportation is clearly a big one but there's certainly other applications. When you look at agriculture — in fact autonomy has been there longer than any other industry. The key thing is going to be cheap and ubiquitous autonomy — this will enable smaller simpler machines to take advantage of the ripple effects of this technology at scale. That could transform how agriculture is done and the cost of harvest and planting and all this other processes that they have to deal with. I think you'll see impacts in insurance for sure. That industry — they're trying to figure out how it's going to work still. There’s obvious other one — mining, construction anything that involved a wheeled vehicle today.
And then, the other one that I think will be very interesting is in basically entertainment. In reality, when you free up a lot of people's times, we like to think that we could use it for education— novel things — but realistically, most people will use that time for entertainment. So now, you've created an entirely new platform for people to consume content. And that's another really interesting thing that's emerging, but that's why companies like Google are so interested in this space.
How do you see the industry as being ripe for disruption today?
Well, I think a lot of things are at stake. Automotive as a whole is a massive industry — something like a third of the US population is employed doing something with a vehicle whether it's being a truck driver or a pizza delivery person or a taxi driver — any number of things. So there's no question in my mind that this is a massive opportunity. When there's a massive opportunity, there's going to be a lot of innovation that is going to come with that. Companies like ours are trying to find the right piece that is important and universal in that whole ecosystem.
When you look at why, the existing players in the space are faced with problems that they are traditionally not good at solving — big car companies don't write good software — we all experience the subpar touch screen dash systems of our car and they're all you know pretty terrible. And now they're going to work on probably the hardest software problem ever proposed by mankind? — it's a bit of a mismatch. I think that it creates a lot of opportunity and car companies are recognizing that and trying to do the best they can with integrating and partnering and all kinds of different things.
On the team, we've built now two autonomous vehicles for the DARPA challenges and now another one for this company. The thing that we saw is when you build these frameworks, there's kind of a foundation that everything is built on and that foundation is knowing where you are.
What is the current issue in the industry that you are helping to address and how are you doing so?
We are solving the problem of localization and what that means quite simply is figuring out where you are on a map. And the key reason why that's important is if you think about a car — it not only needs to be able to stay in a lane, but it has to know information on the road. What we do is we connect the dots for the car. We tell exactly where in this virtual environment that the actual car is located so they can use this precoded information in the map to help make smarter, safer decisions.
Could you tell us a little bit about why you chose to target this particular problem?
We did think about building a full self driving car. But the reality is that that in itself is a massive problem and a lot of people are already focusing on the high level — how do I perceive the environment? how do i detect cars? etc. But what we saw that was falling by the wayside was how do i reliably know where I am in the world and do that at scale. And we saw that as being a massive opportunity, because people were skipping that problem to get to the next problems.
It happened to be something that we all had a ton of experience in as well. Between my team, we have over 50 years of GPS experience. And it's the type of thing where we've built localization systems for the government and the military and for all kinds of different rovers. We have a lot of experience telling things where they are. And I think that's an incredible opportunity for us, because when you look back to the universal solutions — things that all companies need—, I think location. Localization is something that you really need.