Musk said that he is developing pure vision technology for autonomous driving, what is the safety pe

Before the technology was developed, no one knew how safe it was.

In a way, when we mention Musk or Tesla, consumers can't always think in terms of safety. There is no need for everyone to be led by the nose, and always use safety to evaluate Tesla's or Musk's every move. Tesla's self-driving technology does have certain flaws, but this flaw not only exists in Tesla's vehicles, but also in all new energy vehicles and cars with self-driving technology.

1. What's going on?

At the World Internet Conference, Musk took the stage and made a speech. Musk said that he is developing autonomous driving technology under pure vision technology, and realizes autonomous driving through network control of visual nerves. At the same time, a new era of self-driving is coming, and Tesla will also specialize in developing special chips with relevant performance.

Second, this technology has not been implemented, so we don't know how safe this technology is.

As I said above, this is a completely new branching technology. At present, autonomous driving technology has not achieved 100% safety, and there is still a lot of room for development of autonomous driving technology under the optic nerve. In a sense, we need to be tolerant of science and technology, and no science and technology can be 100% perfect at the beginning of its existence.

3. I'm very much looking forward to Tesla's technology.

Tesla is currently the leader of new energy vehicles and autonomous driving technology, and I think it is a very meaningless behavior if consumers keep struggling with Tesla's driving safety issues. It's true that the current self-driving technology is not perfect, and this is also present in other car brands. On the contrary, I think we need to focus on the development and innovation of new technologies, and we are very much looking forward to what changes they can bring to us.

The safety of this technology should be very good, but now it is also because it has just been developed, so there are few people who experience it, and it is currently being improved.

The safety performance of this technology is still very passable, and it has passed the corresponding tests, and it looks really safe.

The safety performance is very good, it is a very guaranteed technology experiment, after the success of the automatic driving does not need us to manually manipulate, but also to ensure our safety.

It's very safe, sometimes people will be sluggish, and they will be more agile by visual control.

In fact, there has always been a big problem with self-driving car technology, and it has also attracted a lot of attention, many people have been questioning this technology, and also believe that this technology is likely to bring greater safety risks. There are many people who don't believe in this technology and don't want to apply it, so the scope of application of this technology is relatively narrow.

There are companies that are not only able to develop self-driving car technology, but also more advanced technologies that may be able to bring us more convenience. Tesla CEO Elon Musk said that he is developing pure vision technology for autonomous driving, when will this technology be successful? I think it will take at least a year for this technology to succeed, and there are three people who say that:

First, the research and development of this technology is relatively difficult.

In fact, the reason why I think it will take at least one year is because I think it will be more difficult to develop this technology. Although Musk was able to make a more detailed plan and analyze the difficulty, I think Musk did not take into account the reality that the technology itself is relatively difficult to develop, so it will take at least a year to develop successfully. <>

Second, the R&D capital required for this technology is relatively large.

If you must develop this technology, you must invest more research and development funds, because the research of this technology must have more investment in the bridge, and only by investing more money can this technology have a faster research and development speed, and the problem of funding is also a problem that Musk needs to solve. <>

Third, the existing R&D technology can not fully meet the needs.

I believe that within one year, this technology cannot be successfully developed, because the existing research and development technology can not meet the needs of the development of this technology, and it may not be able to improve in a short period of time, limited by the existing technology, the research and development of this technology may need to be postponed, and it also needs to wait for the existing technology, more advanced before it can make progress. <>

That's why I think it will take at least a year.

Maybe it will succeed in the near future, but it is very difficult, in fact, Tesla's current self-driving technology is not particularly mature.

I think it will take at least a year for this technology to succeed. Because the research and development of this technology is relatively difficult. There are a lot of R&D resources required for this unbridled travel technology. The existing R&D technology cannot fully meet the needs.

It will take at least 3-5 years depending on the amount of money invested and the technology to be developed.

Pure vision is the right way to go.

I don't think there is any logic in saying this, there are already some problems with their car, and it is very irresponsible to say this.

I think it's okay, if you really make a product with super high security, it will be good for everyone, especially for the guarantee of safety.

I think autonomous driving may be the trend of the future, but it is not yet possible with current technology.

If a large number of self-driving cars on the road together, safety will indeed be improved, each car system networked, keep a distance from each other, and at the same time maintain a formation, front and rear left and right are the observation range, when an accident occurs if a child runs out, all the cars in the vicinity do not disturb whether they see it, socks can be controlled at the same time parking, of course, it will be very safe, provided that there can be no artificial driving vehicles mixed in.

According to Musk's speech, Tesla will completely solve the problem of computer vision algorithms at the end of 2017, that is, before its cross-American self-driving technology demonstration, and verify the reliability of its system in the next year or so, and then provide Level 5 autonomous driving functions for all Tesla models around 2019.

I think it should be able to improve this aspect.

Tesla also pointed out that this feature needs to be built on the premise of the legality verification of the self-driving software system and the relevant policy permission.

I don't think this driverless will be completely comprehensive.

It completely overturns people's perception of unmanned driving.

That is, before the demonstration of its autonomous driving technology across the United States, the computer vision algorithm problem will be completely solved, and the reliability of the system will be verified in the next year or so.

The madness lies in its ability to solve problems related to computer vision perception in advance.

Computer perception does not have manual driving safety, and there are always some blind spots that cannot be perceived.

The vehicle can detect the surrounding environment in 360 degrees through these eight cameras, and the radar is responsible for deeply detecting the distance and travel speed of obstacles ahead, and will not be affected by weather factors such as precipitation, dust, and fog. This almost replaces lidar sensors functionally, while reducing the cost of automation systems and avoiding the additional weight and aesthetics of the vehicle that come with redundant sensor embedding.

Through the release of this autonomous driving chip, Musk once again educated the industry: it is entirely possible for vision-based solutions to achieve L4 autonomous driving.