In this Q/A session with David Liu, founder and CEO of portfolio company Plus.AI, David shares his perspectives and outlook on the future of autonomous vehicles. Plus.AI is developing autonomous driving solutions powered by deep learning.
Why should we be excited about autonomous vehicles?
It's going to change the world and save lives.
What does the future of autonomous vehicle look like to you?
Well it's going to go in stages. In a couple of decades, what you're going to see will mostly be cars driving on their own. When you think about transportation, you think about going some place or going from place to place. You're not thinking about driving. Vehicles become just a tool. It's going to take some time to get there.
In the process, autonomous vehicles are going to take different forms. So today, you look at the the different levels of autonomy going from level 2 to level 5. Most of the vehicles or at least the high end vehicles have level 2 autonomous vehicle driving function — this is more of a safety feature. At some point, we will get to level 3 driving — we will still have drivers, but the machine will do a lot of work for the driver so the driving experience can be a lot less labor intensive and will be safer.
What are the biggest hurdles to autonomous vehicles becoming mainstream?
The first hurdle is technology. People have been working on this technology for the past two decades — it’s progressed a lot especially in the past 5-10 years, but still we're not there yet. Technology is the largest hurdle.
The second hurdle is infrastructure mainly because the road conditions are very different and infrastructure will be optimized eventually for autonomy.
The third hurdle is probably on the regulation side. On the society basis, people need to get comfortable with driving alongside machines, and there are a lot of questions that are unanswered regarding accidents. Even today, for the regulation side, when you use an autopilot or cruise function on the highway, if you get in an accident, is it the drivers fault or is it the manufacturers fault? Those are some of the questions that need to be answered.
Apart from transportation, how do you see autonomous vehicles disrupting other industries?
I think it's going to completely change all other industries. Transportation — that’s the movement of humans on vehicles. When you get into moving matters, moving goods — now you get into logistics; the logistics industry will be revamped and if you change how goods are moved, you're going to eventually touch upon how goods are manufactured and where they're going to be manufacturers. It changes how commerce will be done.There's all these discussions on manufacturing for consumers — personalized orders directly sent to a factory, having the good made on a personal basis and having that shipped to the consumer’s home. That will all become possible if you can make the efficient transportation and logistics system on a magnitude larger than it is today. How society functions will be completely different.
How will infrastructure need to adapt? Are there certain road structures/layouts that would best support full autonomy?
Yeah, I think before we get to complete autonomous driving, the infrastructure needs to change accordingly. If you think about today's highway system and road system and how they are different from a hundred or two hundred years ago — when they only had carriages and horses running around, it's very different. And similarly, we will come to a time when cars on the road are self driving, and I can imagine the roads will be much different — how we're going to design the intersection, how we're going to design the road sign, etc. A lot of that will be different from how it is today. Roads are made for human driving today. For autonomous driving, it will be different and optimized for that purpose particularly.
What is the current issue in the industry that you are helping to address and how are you doing so?
On one hand it's a technology problem. So to get to autonomous driving, there's a lot of problems to solve on the entire chain from conception to localization to execution. Among that series to providing the end solution, we need to do a few things to demonstrate and understand how the technology will evolve. The first order of business is to understand that the technology champions a broad and orthogonal effect on the space. We need to figure out how we’re going to make the world better today, and in the next year — not just a decade from now. So we’re currently figuring out the problems that we can solve and help with today and apply our technology there.