Xpeng CEO examines the current ups and downs of Xpeng self-driving technology.

Every car and their cousin is going to drive us forward መሠረት according to their press sections. In fact, I think there are only a few deep-seated automobiles in the house that focus on amazing ingenuity to reach that ultimate goal. In a short time, that kind of leadership will enable them to have the best ADAS (Advanced Driver Assistance System) features. In addition to Tesla, the largest company I have been following in this field is Xpeng. I know that they have a strong internal team focused on the development of self-driving technology, and I know that they are taking an exciting approach that integrates more types of hardware (LEDs, radar and cameras) than Tesla. In terms of software, I guess there are more software developers that focus on this than anyone else. But In short, I think Tesla is one of the top companies in this field that could transform itself into a pole – not for the time being but in the long run. i am not. Prediction That (I’m not making such predictions for any car, independent start-up or automotive supplier) but it’s an obvious opportunity.

With that context in mind, I was thrilled to see how Xpeng CEO He Xiaoping responded to Xpeng’s “City NGP” (City Navigated Guided Pilot) in its latest technology test drive. As I watched the video, I was even more impressed by the fact that he was very open and critical about the weaknesses he saw in the technology, and praised some of his superior abilities. As I have in my own car, Tesla FSD Beta, I am well aware of the major challenges along the way for such technology, and you can see He Xiaopeng pointing out some of the real challenges in the video. Self-driving software highlights unbalanced issues that prevent a driver from being as smooth and relaxed as usual. Watch the video below and then I will return to some specific things.

By watching this short video of the city’s NGP capabilities, the technology does not look much different from Tesla’s autopilot. It looks great on lane protection and traffic control (with adaptive cruise control), and seems to be very good at switching lanes automatically, but some small wheels are sometimes too active or do not know the context of congestion. – Sealed traffic jams where other cars behave differently and do not give you much space to enter the unlock. Although Xpeng’s Autonomous Driver Center VP Wu Xinzhou said in the video that the future version of the technology will detect such traffic conditions and ask the driver to help him change lanes in those areas. It is not a perfect robotics approach to this test, but it looks right for this level of technology.

He also noticed that Xiaopeng was seen in the Xpeng City NGP Suite when a car entered the line with a slight force, and the software gave some extra space for that car to fit properly. That seemingly natural level of human adaptation is amazing, and it shows how the Xpeng team is working to make urban NPC as smooth and predictable as possible.

Anyone who has followed my comments and reviews of the Tesla FSD Beta knows that one of the main concerns and differences is the pitfalls. No one wants to drill holes, but the system has not yet been able to handle them. He Xiaopeng asked in particular about the wells and how the city’s NGP system would handle them, and Wu Xinzhou pointed out that he still could not understand and avoid them, but that the team was working on that and that the next version of technology would “definitely” know them. Personally, I find it refreshing to be willing to share such weaknesses publicly (not every company emphasizes these things – and certainly not) Via CEO test drive video). That kind of accuracy and honesty is respected, and I think it’s not just me. I also find it encouraging that they focus on solving the problem. I look forward to seeing how well they do with the next semi-autonomous driving technology.

Wu Xinzhou and Lidar spend some time explaining how to help the Xpeng team with things like heavy rain. Again, this Tesla system is 100% based on vision and is a bit of a struggle due to visibility challenges in such rain. I want to compare the two systems side by side in the same rainy conditions! Maybe Xpeng can lend us a car for that one day. 😉

Wu Xinzhou also talks about ways in which the company can help identify and eliminate persistent obstacles. I’m not sure that this is deliberately casting light on Tesla’s path, but Tesla’s vehicles are a known problem. Naturally, Tesla is working on other ways to solve that problem, but it seems interesting to emphasize that this is one of the reasons why they consider Xpeng Leadar as important.

For computer geeks out there, the most impressive note is that Xpeng uses all of this for computer power. “We are the only company in the world that has been able to develop ADAS services at the city level based on 20-30 TOPs of computer power,” said Wu Xinzhou. But instead of touching his own back for that truly amazing success, he immediately went into some self-criticism. But we also have limitations. Computer skills are not sufficient for certain situations. He then talks about some of the edge issues they are working on, such as people sleeping, people and people in wheelchairs sitting.

All in all, I am very happy with what seems to be an open and straightforward place in the Xpeng semi-autonomous driving technology. I strongly recommend that you watch it. If you have any additional comments while watching the video, please write in the comments below and let us know.

Do you appreciate CleanTechnica’s original and clean technology news coverage? Consider becoming a member of CleanTechnica, a supporter, technician or ambassador – or Patreon’s guardian.