Shulou(Shulou.com)11/24 Report--

Robotaxi's climb of Mount Everest is probably in the last night before dawn. Because California Silicon Valley has just reached a key milestone:

Open Robotaxi unrestricted operation! 7X24 hours, unlimited area, no one, can be charged.

Unmanned cars in San Francisco are already on an equal footing with human drivers and are no longer subject to any special controls.

The significance of this milestone is obvious: on the one hand, unrestricted field competition has proved the accumulated achievements of self-driving technology in the past, and safety and reliability are ready to be tested. On the other hand, in the future, autopilot and Robotaxi are ushering in a major historical turning point, which is about to produce new business value.

And, after all kinds of twists and turns, the commercialization of self-driving once lagged behind China's Silicon Valley and recaptured the global self-driving center overnight.

In an era of rapid technological iterations, self-driving and self-driving cars were once overshadowed by other new technologies, but no one can deny the fundamental changes in cars, travel, transportation, cities, and even the economy.

Well, under the racing, it can't be ignored.

Silicon Valley's decision to retake the unmanned vehicle center, which has no restrictions on Robotaxi, was brief and forceful, and was finally pronounced by the Public Affairs Committee of the California legislature where Silicon Valley is located:

Google's Waymo and GM's Cruise's Robotaxi are allowed to carry passengers in San Francisco without restrictions.

24 × 7 round-the-clock, unrestricted urban area, unmanned Robotaxi operation that can charge passengers.

However, the release of this license has not been plain sailing. In the process of birth and landing, new things will inevitably have a fierce confrontation with the old ideas.

The hearing on the new resolution passed by California is said to last as long as six hours. Most of the opposition is focused on security and employment security.

The former is because the unmanned cars of Waymo and Cruise did appear to be unable to handle by the algorithm during the test in North America, resulting in traffic jams.

The latter has been controversial since its birth, and there are even many people with vested interests in the old form who constantly obstruct, oppose and call for restrictions on unmanned cars.

But in the end, the California legislature gave the following advice:

This technology has the potential to improve road safety and play a positive role in San Francisco's economy.

For safety, both Waymo and Cruise operate Robotaxi with a speed limit of no more than 65 miles per hour, about 100km / h, while in urban areas, the speed limit is 35 mph (60km).

In addition, according to official statistics from the San Francisco Municipal Transportation Department, driverless cars operated by Waymo and Cruise have had a total of 12 accidents in the past six months, causing a total of less than 90 points of traffic jams.

This figure is certainly much prettier than human drivers who stumble on the street and suffer from road rage.

Robotaxi, which has no one at all, has not reported any accidents that have caused casualties.

That is why California officials believe that liberalizing restrictions on unmanned cars is good for the economy in the long run.

Because from the perspective of the whole social and economic operation, the popularity of unmanned vehicles is not a zero-sum game, and the unemployment problem will not exceed the normal bearing scope of the society. The improvement of traffic efficiency by unmanned vehicles will improve the overall GDP, which indirectly creates more value and opportunities.

Of course, North American regulators have also left a way back, and the testing and commercialization of unmanned cars is not a "release" attitude regardless of questions:

After releasing the restrictions, depending on the operation situation, we reserve the right to withdraw the resolution at any time.

Similar to the "reinforcement learning" in the AI algorithm, when unmanned cars do a good job and improve their safety, the policy encourages openness and narrows the scope of permission if problems are exposed.

What is the significance of letting go of unmanned cars? compared with the past, the biggest feature of this new development is that there are "no restrictions" on unmanned cars.

Foreign media Techcrunch directly commented like this: a win for the autonomous vehicle industry, the victory of the self-driving industry.

Waymo and Cruise said it was "the real starting point for business operations".

Normally, the testing and deployment of unmanned vehicles on the road need to be approved by the California Motor vehicle Administration (DMV), and the California Public Utilities Commission (CPUC) is required to approve the charging of unmanned vehicles to passengers.

Before the latest decision, the landing of the two companies in San Francisco was not fast.

In fact, Waymo started free internal testing in San Francisco as early as 2021, and it is still at the stage of providing free services.

Cruise suffers from restrictions on fees. The fee is only allowed in a small part of San Francisco, and the time is limited to 10:00 to 5: 30 a.m.

Therefore, considering from the technical level of L4, San Francisco's full liberalization of unmanned cars is not only a preliminary recognition of technical safety and reliability, and the opening of more large-scale and formalized testing, but also the lifting of every restriction before. Is of extraordinary significance.

First of all, there is no time limit, which is equivalent to activating the hidden value of the unmanned car. At the data level, the system can collect data under various weather, time or lighting conditions, cover more long-tail scenes, and improve reliability.

In addition, the unmanned vehicle 7X24 hour operation, can be more reasonable arrangement of "attendance-charging" plan, so that the vehicle idle, maintenance time to a minimum, efficiency than human-driven ride-hailing, the cost is reduced to a minimum.

The urban area is unlimited, and it is also a source of rich data, covering more long-tail scenarios, and the most important improvement for the system lies in "generalization", that is, the ability of the same set of basic algorithms to deal with different cities and different road conditions, and the generalization ability is good enough to deploy Robotaxi services to more cities more quickly and reduce development costs.

The unmanned main driver and the allowed fees are put together, which completely clears the obstacles to the landing of the high-level self-driving business above L4.

Robotaxi, or all high-level self-driving landing, including self-driving trucks, minibuses, etc., the biggest obstacle is the high cost, unable to form business logic.

This is reflected in the self-driving vehicle itself, which needs to overlay the self-driving kit on the basis of the "car", including sensor hardware and the sharing of the cost of AI algorithm development. Of course, it is possible to reduce the cost of vehicles by partnering with car companies for pre-loading production, but this way is still being explored in the industry, especially in North America.

So the problem with the commercialization of unmanned cars before is that a self-driving vehicle that is much more expensive than an ordinary ride-hailing car + a security guard whose wages are the same as those of human drivers. However, the services provided are unlikely to be more expensive than online ride-hailing, or even free at this stage.

Therefore, if there are security officers on Robotaxi, it will never be possible to achieve commercial landing.

Unmanned driving plus permissible fees not only clears the way for the smooth birth and maturity of the Robotaxi species, but also injects confidence into the self-driving industry, giving hope to players who have been technically committed for more than 10 years.

So summarize the significance of the new progress, not only declare that L4 technology has become mature, safety can be tested, but also prepare the prerequisite for the next stage of high-level autopilot scale and commercialization, paving roads and bridges.

The global self-driving car racing pattern changes overnight that Silicon Valley has "regained" its position as the center of unmanned cars overnight because North America has lagged behind China in L4 commercialization for at least the past 2-3 years.

For a long time, Waymo and Cruise, the two most powerful head players, can only be tested in limited areas designated by San Francisco and Phoenix. There are either mandatory requirements for security guards or restrictions on fees.

The "cold spell" that began on the North American L4 track last year was marked by the collapse of the Argo AI leaning against Ford and Volkswagen, as well as various personnel and public opinion earthquakes affecting Waymo and Cruise.

In fact, the source of these problems can be traced back to the wait-and-see, hesitation of North American regulators and the slow opening of policies, which are variables that have been "delayed".

After all, autopilot is an extremely expensive project, as big as Google, can not use love to generate electricity.

Official attitudes also affect the acceptance of self-driving by ordinary users. For example, there are more and more fierce anti-driverless groups in North America, often taking to the streets to stop Robotaxi with physical bodies, making the otherwise safe and normal operation of the scene dangerous.

In the past two years, China has made progress that is the envy of North America. Beijing, Guangzhou, Shanghai, Wuhan, Yangquan, Chongqing, Changsha and other regions have successively introduced various policies and demonstration areas to encourage driverless cars.

For example, Beijing, the first to issue Robotaxi test license, the first to open the test operation with safety guards, and then the first to allow the main driver unmanned, and open fee test operation.

China takes the lead in the first stage of "experimental verification" of L4.

The next stage of "scale" and "commercialization", Silicon Valley has taken a big step, skipping the domestic "step by step" process, one step in place, suddenly ahead of China.

However, from a longer-term perspective, domestic L4 still has a unique comparative advantage.

From a technical point of view, China's road challenges are more complex, but it also represents a higher upper limit of system capacity. With the same power base and team size, the high-value data collected by Robotaxi running in China at the same stage is higher than that in North America.

But the most important thing is China's perfect industrial chain, so that the cost conditions for commercial landing are much better than those in the north.

For example, lidar, domestic players have been able to make semi-solid products thousands of yuan each, and can perfectly support the pre-installed mass production of the main engine factory.

So at the level of cost reduction, Chinese players can quickly reduce vehicle costs to the same level as online ride-hailing in addition to removing security guards. For example, the cost of the sixth generation Robotaxi, which Baidu Apollo will launch at the end of the year, has been reduced to 250000.

The technical basis and industrial chain conditions of China's unmanned vehicles are no longer a problem, everything is ready, only policies are in place.

The pattern and center of the global unmanned car is still facing a reversal.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.