00:00China is to accelerate its push for the country's tech and science sectors to become self-reliant,
00:05boosting research and development within the country.
00:08CGTN's Wang Tianyu spoke with James Pung,
00:11the founder and CEO of the Chinese autonomous vehicle technology company Pony AI.
00:17From a software side or from AI side, we have developed, we call the Pony World,
00:24which really is a world model where we let the vehicles to practice by itself.
00:30So essentially, you can think of this as a player who practices in a virtual world for so many, many
00:38times.
00:39And then in real world, it's more like a muscle memory.
00:42So that's from a software side.
00:44And from the hardware side, especially the vehicle side, we actually have multi-layers of redundancy
00:50because we know in real world, any hardware may potentially fail.
00:55And in any failure situation, we need a redundancy to kick in, to take over.
01:00So from a census point of view, we have 360 degree coverage.
01:05And at any single point, our requirement is we should have at least two sensors to look at it.
01:12And then from a system point of view, we have three layers of redundancy.
01:17Main system, we have a redundant system that can allow the vehicles to pull over safely on the side of
01:23the road.
01:24And even on the third layer, we have a very simple system that allow us to stop in lane.
01:32All right.
01:32So your car is also running overseas markets like Europe and Southeast Asia.
01:37What adjustment to strategy do you think are necessary to succeed in those areas?
01:43Two important reasons.
01:45One is related to regulation.
01:47The second is with cultural or the local environment.
01:50From a regulatory point of view, in foreign countries, it has a lot to do with the local traffic regulations,
01:58has a lot to do with data security, with the privacy concern and all that.
02:03And for the second is related to the local traffic pattern.
02:08Essentially, the local culture where how people take the transportation options.
02:13You can think of us as a complement to the existing taxi or ride hailing.
02:19So how the taxi and ride hailing matured in those markets is also important as a reference for our success.
02:27You have partnered with many OEMs and ride hailing platforms instead of just one.
02:34So I'm curious, why did you choose a broader cooperation strategy instead of, you know, focusing on one model?
02:42Didn't that make it look more unified and cost effective?
02:46Yeah, that's a great question.
02:47From simplification of tech development, it's probably a bit easier to focus on one partner.
02:57But because we are doing the real-world application, there are different markets, right?
03:02There are different cities.
03:04Typically, each city or each market prefers certain brands or certain type of vehicles.
03:11So it's more to do with the market demand rather than the simplicity for technical development.
03:19You have a plan to expand the robot taxi fleet to over 3,000 vehicles by the end of this
03:25year.
03:26What are some scaling challenges that you can foresee right now?
03:30Well, I think one of the challenges is on the supply chain stability, right?
03:38Especially now we see the price of the memory is skyrocketing.
03:44So how we can ensure the supply of all the components and the sensors so that we can manufacture 3
03:51,000 vehicles.
03:52Another challenge, of course, is how we deploy them.
03:55And the deployment relies on the government to issue licenses.
04:01So we need to ensure all the policy and license issuers are in place.
04:10And the process is to ensure the product of the market does not include the flow of all the individuals.
04:10So, if there are some expectations, we can also ensure that the controller is not included.
04:10You
Comments