Recently, during his appearance at the Boao Forum for Asia, Yu Chengdong made a bold statement again: With Huawei entering the autonomous driving sector, the industry has been redefined.
"Our ADS advanced autonomous driving system has been acknowledged by everyone. In terms of the cockpit and autonomous driving experience, we're universally recognized as the best."
Wow, that's quite straightforward! Yu Chengdong also dropped a bombshell: Huawei's Smart Car Unit (BU) is expected to turn a profit this year. Now, that's a big deal considering BU has been bleeding money since its inception. They lost a whopping 10 billion CNY the previous year and another 6 billion CNY last year, making it Huawei's only unprofitable division. If they manage to rake in profits this year, the hero behind this success story undoubtedly has to be the Wanjie Xinyuan M7. In February alone, they sold a total of 21,083 units, ranking third in sales among all models.
So many people are buying it, mainly because of Huawei's advanced intelligent driving features.
The intelligent driving version of the new M7 is priced 20,000 to 40,000 higher than the regular version, but according to official data, more than half of consumers are willing to pay extra for the intelligent driving feature.
After its release, this car has undergone several OTA updates, all of which are closely related to intelligent driving.
Coincidentally, recently, the Wanjie Xinyi M7 has pushed another OTA update, claiming that the urban NCA, which can be driven nationwide, has been fully pushed this time, and all car owners can use it.
It is said that after this update, the experience of urban intelligent driving has been improved again, so we also got a Wanjie Xinyi M7 to see what level it is at now.
First off, this system claims to cover 99% of the country's roads, not just highways and urban arteries, but also rural roads.
We tested it in Hangzhou, running it through elevated roads, tunnels, urban arteries, mixed traffic areas, and even a stretch of village lanes. Basically, we threw everything at it, no matter how complex.
So, it turns out the hype was real. The NCA was usable throughout the entire test, and this version instills a lot more confidence, especially at busy intersections with lots of cars and people. It's noticeably more stable than before.
Let's talk about something fun. Active safety has always been a key focus for Huawei's smart driving technology.
Recently, a clip went viral online showing a vehicle, equipped with Huawei's smart driving system, performing an emergency maneuver at a speed of 130 km/h on a highway.
Yes, that clip. Initially, various automakers claimed it was their car in the video, but it turned out to be the new M7.
This time, there's an update on the active safety features of the AskJie New M7, with the addition of a new side collision avoidance assist. This feature works by pulling you back if you're getting too close to obstacles on the side while you're distracted.
I once had a real-life demonstration while parallel parking.
I was getting too close to the curb, almost scraping it, when the car automatically corrected itself by reversing slightly. Unfortunately, I didn't capture it on video.
Here's an official demonstration clip. Imagine replacing the parking cone with the curb and visualize it yourself.
Even though this wasn't captured, I did manage to test out two aspects of the active safety features.
The first one involves the common scenario of "dooring," where the car detects a vehicle ahead opening its door and then steers to avoid it by swerving.
The second one is about unexpected obstacles. The usual cones and barriers, our smart driving systems can identify them pretty accurately. But sometimes, there are other objects on the road that the machines don't recognize. For instance, if there's a square object up ahead on the road, our car can detect it from about 10 meters away, then it'll shift to the left lane to bypass it before coming back to its original lane.
Alright, after touching on active safety features, let's dive into urban commuting.
Right off the bat, we're putting Huawei's NCA to the test.
We're talking a street where pedestrians are crossing, a bunch of e-bikes ignoring the rules, maintenance vehicles on the side, and ride-share cars that could stop right in front of you any second.
I originally thought the system wouldn't be able to handle such a complex scenario, but it turns out it managed to navigate through it successfully on its own. This achievement has a lot to do with the newly added "Priority for Traffic Efficiency" feature.
When activated, the car spots an obstruction and figures out a way to avoid it. We just so happened to experience this; the car maneuvered around the obstacle and then immediately returned to its lane.
Compared to the previous setting where it would get stuck as soon as it was blocked, this approach is actually closer to how humans handle the situation. However, I'd still like to remind everyone to observe closely and be ready to take over at any moment. When there's someone or a non-motorized vehicle ahead, the car will inch forward, waiting for an opportunity to pass.
A cool feature is that the dashboard display can show whether the car in front is signaling a turn. The system gets this, allowing it to make predictions ahead of time.
The rest of the journey is just smooth sailing on the expressway. The Huawei NCA system performs as reliably as always, smoothly handling lane changes, overtaking, and signaling all on its own.
This time around, not only has the NCA been upgraded, but the smart parking system has also been updated with the addition of Remote Park Assist (RPA) feature. When encountering tight parking spaces, you can exit the vehicle and use the mobile app to remotely guide the car into the spot.
Another nifty feature is that shifting the car into reverse automatically brings up the parking interface, so you don't have to manually navigate to this screen anymore.
In addition to the above, the AskMoon M7 has some practical details upgraded, like the newly added automatic matching of the highest speed limit on roads.
Actually, before, you could also let it cruise at the speed limit, but it often required manual adjustments when switching between different roads. For example, moving from a city road with a 50 speed limit to a freeway with an 80 speed limit, it used to need to detect a speed limit sign before the car would accelerate. But now, once you're on the elevated road, the speed picks up automatically, which feels seamless. Honestly, these past few days, I haven't had to adjust the speed myself at all.
Also, there's a hands-off-the-steering-wheel detection feature that now has two settings. The standard setting alerts you after 15 seconds of hands-off, while the comfortable setting alerts you after 40 seconds. For beginners, I'd recommend starting with the standard setting to get used to it. Of course, for seasoned drivers like me, the comfortable setting is the way to go.
Alright, that about wraps up the updates on the new M7 smart driving tech from Wenjie. From what we've seen, Wenjie's level of autonomous driving tech is seriously impressive, easing the burden on drivers during everyday commutes. Even our photographer, who's a huge fan of performance gasoline cars, couldn't help but give a thumbs up after shooting the Wenjie, practically itching to place an order on the spot.
To be honest, when the new M7 Smart Driving Edition from Wen Jie was released last year, its NCA feature was pretty handy right off the bat. Now, after several updates, it’s become even smoother, which probably doesn't come as a surprise to many.
But before the new M7 came onto the scene, Huawei's presence in the smart driving arena wasn't really that loud.
After all, Huawei's automotive BU was only established in 2019, giving them just 5 years in the game, while the folks at companies like NIO, XPeng, and Li Auto have been at it for nearly a decade.
So, how did Huawei manage to stand out? This time, we also got in touch with some insiders at Huawei to dig into the story behind their success.
First off, while the automotive BU was a later addition, Huawei's R&D in automotive technology actually kicked off back in 2012.
At that time, the concept of autonomous driving wasn't really a thing yet. They established a "Connected Vehicle Lab" under the "2012 Lab," because it involved communication technology, which is Huawei's forte. Not long after, when autonomous driving started gaining traction in the industry, Huawei was quick to enter the scene. Once they got involved, they realized that many of the technologies required for autonomous driving were actually in line with the ICT technologies they had been working on for decades. For example, lidar involves using lasers to scan objects, which is related to optical communication technology.
Huawei has a dedicated team for optoelectronic technology research, and they're diving into laser radar, kind of like how a League of Legends player switches to playing pesticide—maybe not the most familiar with the hero's skills, but definitely adept at last-hitting, bush camping, and teamfight awareness, so picking it up is a breeze.
Similar to autonomous driving, which involves chips, sensors, algorithm models, and more, Huawei has its fingers in all those pies, meaning it's got a solid foundation to start with.
But having a good foundation is only part of the equation; to make strides in smart driving, you also need to pour in a ton of manpower and cash.
From the establishment of its Automotive Business Unit to the end of 2022, over roughly four years, Huawei has amassed a research and development team of 7,000 people and invested over 20 billion RMB in autonomous driving and related technologies.
With the resources and talent in place, how do they make it work? Huawei folks say they approach smart driving much like they do wireless communication—"commercialize one generation while researching the next." In other words, they're already brainstorming the next-gen tech while the current one is just hitting the market.
Before the 2021 Shanghai Auto Show, a video showcasing Huawei's smart driving capabilities with a Gigafox vehicle went viral across the internet, giving everyone a taste of what Huawei's smart driving tech can do.
However, at that time, the plan was "Map-based NCA," meaning the intelligent driving system required the assistance of high-definition maps to recognize roads. However, while developing "Map-based NCA," Huawei was already researching "Map-free NCA," which is the solution we are currently using.
Both technological paths were developed simultaneously, and whichever met the conditions for commercial use would be rolled out to consumers first.
Similarly, in the development of both Highway NCA and Urban NCA functions, Huawei pursued both paths concurrently. However, because highway driving is simpler, it was released first. Some automotive companies prioritize developing highway intelligent driving before tackling urban driving.
With significant investment, Huawei's internal model iteration speed is rapid, with updates typically occurring every 5 days, and this iteration speed is unaffected by sales volume. Of course, the more people using NCA on the road, the better, as more data will help optimize the models further.
All these factors combined, us consumers can truly see the remarkable progress in Huawei's intelligent driving technologies right off the bat. The massive success of the Wenjie new M7 has also shifted a lot of consumers from not understanding intelligent driving to gradually getting to grips with it. At this stage, whether a car's intelligent driving capabilities are up to snuff can actually influence their buying decisions. It's not enough for intelligent driving to just work; it needs to work well.
That's why, since its launch, the Wenjie new M7 has been rolling out an OTA update pretty much every month. Through continuous updates, it's making intelligent driving more and more user-friendly.
But, all that said, even with Huawei's intelligent driving being so powerful, I still want to add: it's merely an assisted driving feature and doesn't reach the level of L4 autonomous driving. It's still in a stage where it needs to learn a lot to identify and address its deficiencies.
For the foreseeable future, co-piloting between humans and machines remains the mainstream.
So, let's not kid ourselves that with Huawei's smart driving, we can completely relax and expect no traffic accidents whatsoever.
As for those videos online showing people using smart driving for extended periods, texting, or even dozing off, let me reiterate, this is highly irresponsible behavior!
Speaking of which, smart driving is something that's both expensive and technologically demanding. It's precisely because of these high barriers that companies of Huawei's scale manage to pull ahead in the smart driving race.
Through years of technological accumulation, continuous investment, in-house development of both software and hardware, and bolstered by its vast user base, Huawei has achieved a faster pace of evolution than its competitors.
"The best assistant" - that's my fair assessment of Huawei's smart driving.
