In November 2017, Waymo announced the start of an automatic driving test in the absence of a driver. Last week, General Motors requested the federal government to approve production of a driverless car without a steering wheel and brake pedal, and plans to launch it in 2019. These historical transitions are interpreting one thing: the reality that drone is moving from the laboratory to the commercialization of products.
Recently, a new report from consulting firm NavigantResearch attracted the attention of the automotive community. This report lists and ranks 19 representative driverless car companies.
The specific ranking is fully displayed in the following chart. The abscissa represents the strategy, and the ordinate represents the degree of execution.
Criteria for judging these players include: goals, market strategies, partners, production strategies, technology, sales, product quality and portfolio, market share and reliability, and more.
* 2017 ranking of 19 autonomous driving companies
The 19 companies are divided into four categories: leaders, competitors, challengers, and followers. The following is a list of what these four types of companies specifically include (ranked in sequence).
Leaders: General Motors, Ford, Waymo, Volkswagen, Daimler-Bosch, Renault-Nissan Alliance, Ambow, BMW-Intel-FCA
Competitors: Volvo - Autoliv - Ericsson - Zenuity, PSA, Navya, Baidu - BAIC, Jaguar Land Rover, Toyota, Hyundai
Challengers: Uber, Tesla, Honda, and Apple
Followers: None
Top 10 in the overall ranking: GM, Waymo, Daimler-Bosch, Ford, Volkswagen, BMW-Intel-FCA, Ambow, Renault-Nissan Alliance, Volvo-Autoliv-Ericsson-Zenuity, PSA.
To dominate the driverless car industry, in addition to advanced autonomous driving technology, it is also necessary to have the ability to mass-produce cars, equipped with the necessary sensors and computing hardware. In this regard, Silicon Valley technology companies and OEMs face the opposite challenge. Waymo has always been a leader in driverless software, but it needs to find a partner to help it build the car that runs the software. On the contrary, the car company knows how to make a car, but it does not necessarily have the expertise to make the complex software needed to fully automate a car.
To solve this problem, General Motors acquired a Silicon Valley company called Cruise two years ago. At the same time, Waymo has reached a small deal with Fiat Chrysler for 600 cars, but is still looking for a long-term partner, which will help expand the planned driverless taxi service.
Tesla, as the only Silicon Valley company with extensive automotive manufacturing experience, should have the advantage of making full use of drones. But this is not the case. Among the 19 companies, Navigant Research ranked Tesla and Apple in the final.
Breaking up with Mobileye, Tesla is struggling
According to the understanding of new drivers, Tesla launched Autopilot with driving assistance in 2015, which was a leader in the field of automated driving. The initial version of the autopilot used the hardware provided by Israeli company Mobileye, and a fatal autopilot accident in 2016 caused the two companies to part ways. Since then, Tesla has been working hard to build its own autopilot system, the second edition of Autopilot. Tesla CEOElonMusk stated that it is expected that full autopilot will be implemented in 2019.
If Musk can achieve this schedule, Tesla will undoubtedly become one of the leading companies in the driverless car race, but at present, Tesla's development has not been smooth.
NavigantResearch wrote in the report: “Since the first launch of Autopilot at the end of 2015, Tesla’s autopilot system has stalled or even regressed in product applications.” “More than a year after the launch of the second version, Auto The pilot still lacks some of the original features, and the user feedback system has a lot of unpredictable behavior."
"In the TED talk in May 2017, Musk stated that it was building a Level 5 autopilot system." In other words, before 2019, only a software update will be fully automated. In response, NavigantResearch made a direct conclusion. "This is unrealistic."
NavigantResearch pointed out that the reason is that Tesla's current system lacks several key components needed for full autopilot, such as keeping the sensor clean and unobstructed in bad weather, and most of the safety redundant systems.
On the other hand, both Waymo and General Motors are showing off their prototype unmanned cars with a lot of security redundancy. Both companies have cars with at least two on-board computers. If one of them is broken, the other will take over control. They also have redundant power supplies and redundant steering and braking to ensure that vehicles can safely park in the event of a single component failure.
Today's Tesla cars also lack a lidar sensor, and most companies in the industry consider this part a necessary condition for fully autonomous driving. "Even Nvidia has expressed skepticism whether Tesla's computing hardware can reliably support full autonomous driving," said Navigant Research.
Therefore, it is highly doubtful that Tesla promised to upgrade to fully automatic driving by 2019.
But the bigger problem is that Tesla does not seem to have made rapid progress like other companies in the industry. The extremely slow introduction of Model3 has consumed most of Tesla's energy. In the fourth quarter of 2017, the Model 3 delivered only a negligible 1,550 cars. Last year, Tesla’s automated driving department lost some key engineers and managers.
Musk has always been able to finally honor his goal, even if it was later than the original scheduled time months or even years, so the possibility of Tesla eventually mastered completely unmanned technology exists. The problem is that other industries are growing faster than most people expected a few years ago. When Tesla finally achieves full autonomy, it may find that the market is almost saturated.
Universal acquisition of Cruise returns
In Navigant Research's assessment report, GM is considered to be the industry leader of driverless cars up to now, but we can easily see why. Two years ago, the car giant acquired Cruise, a little-known driverless car company. Since then, Cruise has been thriving.
Usually this type of acquisition has certain risks because the bureaucracy of the parent company of the new company will stifle the innovation culture of the new company, which is the key to the success of the startup. Fortunately, GM has successfully given CruiseCEOKyleVogt enough autonomy and influence to seamlessly improve the technology of driverless cars.
Therefore, Cruise has become one of Waymo's biggest competitors in terms of completely unmanned technology. Although Waymo has transferred most of its testing work to the regulated city of Phoenix, Cruise focused on testing in San Francisco. Vogt believes that San Francisco has a better test environment, and cars can encounter unusual and difficult situations, such as construction areas, pedestrians and emergency vehicles, which will enable Cruise to improve software faster.
However, Cruise's biggest advantage is that as part of General Motors, it can easily get the car design and the underlying architecture, which is not available to any technology startup company. Cruise's engineers have been working closely with general engineers to design a modified ChevyBolt for unmanned driving. Last week, General Motors requested approval from the federal government to begin production of a modified Bolt without a steering wheel or pedal next year.
Once Cruise is convinced that its software is safe and secure, GM can start producing these vehicles. This means that even if Cruise is not the first listed company (it looks like Waymo will be listed first), it will also be able to scale up faster and eventually occupy most of the market share.
Waymo has the best technology and now only needs to produce a lot of cars
As a company that takes seriously unmanned driving technology, Google has always been a leader in software development for driverless cars. In November last year, Google’s driverless car project Waymo announced that it will start testing unmanned cars. This shows that they have confidence in the safety of their own technology. At present, Waymo has not announced the date of commencement of commercial operations, but it has already implied that it will only take a few months, not a few years.
But the biggest challenge for Waymo was after it launched a driverless taxi service (most likely in the Phoenix area). Waymo's business model is to provide driverless taxi services. The vehicle is owned and operated by oneself, but it does not have any experience in manufacturing cars.
Manufacturing a car is a complicated process that requires a lot of money and labor. Waymo knows he will not master the automotive industry and is always looking for a partner. According to the new wisdom driver, Waymo has signed several small agreements with Fiat Chrysler to provide Waymo's proprietary sensors and built-in hardware for 600 Pacifica minivans. However, Waymo needs more than 600 cars to conduct a full-scale taxi service in Phoenix, not to mention all parts of the country and other cities in the world.
NavigantResearch said in the report: "With the end of 2017, Waymo still lacks a broader long-term manufacturing agreement with automakers to continue to provide one or more automotive platforms." And this, "It's still Waymo's success The stumbling block."
This is partly due to the inconsistent motivation between Waymo and its potential automotive manufacturer partners. Waymo's business model is to provide a Waymo-branded driverless taxi service that may cut off the connection between the car company and its customers. Automakers understand that if Waymo succeeds, the auto company may become a commodity hardware supplier.
In addition, the manufacturing of automobiles is a capital-intensive business. According to the understanding of new wisdom, Waymo has cooperated with Ford in 2016. However, the cooperation eventually collapsed because Waymo did not want to put some of Ford's costs on the front to expand production capacity. Thousands or even millions of electric light passenger cars carrying Google software.
For Ford, the car is a low-margin and high-risk business. If Ford's services are impacted, they will face huge risks, and Waymo can get most of the profits. If engineers work for different companies, it's even harder for engineers to work together because they worry about unwittingly revealing trade secrets. Cruise and General Motors engineers don't have this problem because they are all from the same company.
None of these obstacles can be overcome. Waymo, the parent company of Alphabet, has a lot of money, and Waymo will sooner or later find people to produce cars. However, Waymo faced the danger of not being able to take full advantage of its early technology leadership to quickly scale up.
Other car companies are also making steady progress
Although this acquisition of Cruise made GM ahead of other auto companies, most other auto companies are working hard to develop driverless car technology, and companies such as Daimler, Ford, Volkswagen and BMW are only behind GM.
Daimler plans to launch a model with advanced driver assistance with smart drives in 2018. This feature is called "one of the most competitive models currently available." Daimler has formed a strategic alliance with Tier 1 automotive supplier Bosch to develop fully driverless technology.
Last year, NavigantResearch ranked GM and Ford as top-tier car companies, but in this year's report, Ford dropped several places. Although General Motors has put all of its driverless car chips into Cruise, Ford has also made several different investments in self-driving car technology. Last year, Ford invested $1 billion in a little-known startup company, ArgoAI. In the previous year, Ford announced the establishment of a Silicon Valley subsidiary called Ford Smart Mobile.
Navigant Research wrote: "Ford's overall product rhythm is slightly lower than the company ranked ahead of it." When Ford said in 2016 that plans to launch completely unmanned technology by 2021, sounds like an ambitious goal, but It now looks a bit behind the Waymo, GM and other companies.
Audi's Audi brand has always been a leader in driver assistance technology. Last year, Audi launched a new “traffic jam pilot” that allowed fully automated driving in traffic jams to be less than 37 miles per hour (approximately 59 kilometers). However, the long-term strategy of Volkswagen for fully self-driving cars is still unclear. In the United States, the availability of traffic congestion pilots still requires regulatory approval.
BMW and Intel (Mobileye acquired in March 2017) and Fiat Chrysler formed an alliance to develop driverless car technology. "BMW's semi-autonomous driving system is getting more and more mature," Navigant Research said in the report, but the long-term plan for the transition to fully self-driving cars is still unclear.
Mica Tape,Single-Side Fiber Brown Mica Tape,Tung Oil-BMI-Expoxy Mica Paper
Thermal Expansion Sheet Co., Ltd. , http://www.insulation-sheet.com
没有评论:
发表评论