This article is from WeChat official account:Sixingren (ID: guixingren123)< span class = "text-remarks">, author: Juny, edit: Vicky Xiao, from the title figure: Cruise’s official website
In the fall of 2013, Kyle Vogt, who had just turned 30, walked into the San Francisco YC incubator with Cruise’s project planner, becoming YC’s first autonomous driving project.
Two and a half years later, Cruise, which had fewer than 50 employees, was acquired by GM for US$1 billion, marking the largest acquisition in Silicon Valley in the automotive industry at that time.
On January 20 this year, Microsoft, which has been looking for opportunities to cut into the smart car track, finally chose to join hands with Cruise and announced that the two parties have become long-term strategic partners in the commercialization of autonomous driving.
Behind its establishment less than 8 years ago, Cruise has now assembled the technology giants Microsoft, automobile industry giants GM, Honda, financial giant Softbank and other industry leaders. The total amount of financing exceeds 9.2 billion US dollars, and the latest valuation has exceeded At 30 billion U.S. dollars, in the field of autonomous driving, it is regarded as the second largest in the industry after Waymo.
In fact, there are not many companies in Silicon Valley doing autonomous driving research and development in recent years. Cruise has not sold a car yet. Why does it stand out in the big waves and even compete with Waymo and Tesla? ?
We try to find the answer from the story behind Cruise.
I created Twitch, but I love the machine even more
Behind every legendary company is a genius founder. Cruise is no exception.
Although Kyle Vogt was just 30 years old when he founded Cruise, he has already participated in two successful entrepreneurial ventures and has already achieved wealth freedom.
In 2006, Twitch founders Justin Kan and Emmett Shear came up with the idea of building a live broadcast platform after selling their start-up company KIKO, but they didn’t know how to implement it in camera technology, soThey sent an email looking for “hardware hackers” to the MIT engineering school list server.
A junior who is obsessed with machine manufacturing immediately responded. Justin and Emmett briefly talked to the student about their ideas in the MIT coffee shop, and then flew to San Francisco. When the plane landed, they received an email detailing how to make a live broadcast to meet their needs. The camera scheme. This student is Kyle Vogt.
“This guy is so cool, we must get him in!” The two brought Vogt to Silicon Valley and let him lead Justin.tv’s product development. And Justin. Tv is the predecessor of Twitch, which was later acquired by Amazon for US$970 million and is currently the largest game live broadcast platform in the United States. It is also Kyle Vogt who wrote the first line of code for Twitch.
Twitch founders Justin Kan (left) Kyle Vogt (middle) and Emmett Shear (right)
Since then, Vogt has also participated in the development of Justin Tv’s spin-off project Socialcam. This video sharing application called “Video Instagram” by the outside world has been sold by Autodesk for $60 million after only 8 months of operation. Acquisition.
Although he has achieved great commercial success in emerging software development, what Vogt has been unable to let go is his dream of making machines.
Vogt once said that he was born for machines. At the age of 13, he made a 200-pound robot to participate in BattleBots, the famous robot fighting competition in the United States, and was invited to participate in live competitions twice. While studying for a degree in computer science and electrical engineering at MIT, he developed a single-arm robot for an abandoned safe that he accidentally found downstairs.He cracked the code and organized a group to participate in the first DARPA driverless competition, which was regarded as the starting point for autonomous driving research in the 21st century.
Even after co-founding Justin.tv, Vogt still did not forget his machine dream while coding the code, and tried to develop a beer delivery robot plan for the company.
The reason why he finally chose autonomous driving was the inspiration of his 14-year-old road trip with his father.
“At that time, my father drove me from Kansas to Las Vegas to participate in a remote-controlled combat robot competition. The journey was very boring. I thought at the time, in fact, I only need to read the lane markings and keep the wheels stable. Done.” Vogt said in an interview.
And 16 years later, his epiphany became his new mission. In 2013, Vogt finally made up his mind to leave Twitch, which was in the limelight at the time, and entered the YC incubator from scratch with the autonomous driving project Cruise.
YC has not been involved in the incubation of projects in the automotive field before, and Cruise is their first in the industry. And the reason for choosing Cruise is largely because of Vogt’s strong personal talent and ability.
YC founder Sam Altman once said that Vogt is the only entrepreneur he has ever seen who can make a completely blank project into a complete software and hardware prototype within three months, and has a strong technical ability At the same time, he can clearly recognize the market, brand and consumer experience, and think that Vogt is a very rare genius.
And Vogt did not live up to YC’s expectations. Less than a year after the establishment of Cruise, the team took out the first self-driving cruise kit RP-1, and took YC’s partners to the Silicon Valley 101 Highway. Wandered around. Although the products at the time were not very mature from time to time, investors were still excited by the effects and the possibility of delivery. AltmanImmediately decided to increase investment.
In September of the same year, Vogt also hired Andrew Gray, the former chief engineer of Tesla’s autopilot software for Cruise, as the company’s vice president, which caused a lot of splash in the smart car industry at the time.
Taking Waymo as a catch-up goal
In Vogt’s view, The best one currently doing in the field of autonomous driving is Google’s Waymo. Google’s many explorations in the field of autonomous driving have also saved a lot of research time for latecomers. When he founded Cruise, he also took Waymo as the main competitor to catch up and in the future.
But the development of Cruise has not been smooth sailing.
In the early days of its establishment, Cruise aimed at the C-end market, which produced kits for modified vehicles and sold them to consumers, allowing ordinary vehicles to have limited autonomous driving capabilities, but the market generally took a wait-and-see attitude towards autonomous driving technology. Under the circumstances, the performance development is not satisfactory.
Starting in 2014, Vogt decided to change its strategy, shift its business to the B-end market, and began to focus on providing software development services for unmanned vehicles for automakers. Since then, Cruise has entered the fast lane of rapid development and large amounts of financing.
In June 2015, Cruise first obtained a road test license for unmanned vehicles issued by the California Department of Motor Vehicles. A few months later, veteran car manufacturer General Motors bought Cruise for US$1 billion. This was a Silicon Valley automobile at the time. The largest acquisition of the year also made Cruise, which has only been established for more than two years, remained in the limelight.
And GM’s attitude towards Cruise is also very special. Basically, Cruise is “let go” as an independent department. Not only does Cruise keep its headquarters in San Francisco, it also gives Cruise great discretion in business. The acquisition process of GM is very rare for established companies like GM.
Mary Barra, president of GM, said at the time that the traditional automotive industry should learn how to cooperate with Silicon Valley technology companies, and does not want to stifle the innovation genes of Silicon Valley companies through technological control and enterprise integration.
After joining GM, Cruise rapidly expanded the company’s scale to thousands of people, acquired two technology and chip companies, and began to substantially launch road testing and simulation training for unmanned vehicles.
In addition to developing an end-to-end 3D engine environment called “Matrix”(The Matrix), which runs 30,000 virtual instances every day In the actual road test, Cruise chose a different test route from Waymo’s “rural surrounding the city”, decided to “first bite the bones” and put the main road test routes in densely populated areas such as downtown San Francisco and downtown Detroit.
Starting in 2017, all aspects of Cruise’s test data have increased significantly.
In 2016, Cruise’s road test mileage was only 10,000 miles, and in 2017 it increased exponentially to 125,000 miles. At the same time, the frequency of manual intervention required for every thousand miles of the Cruise self-driving car has also achieved an order of magnitude leap from 18.5 times per thousand miles to 0.84 times per thousand miles. Although this number is far lower than Waymo’s 0.18 times/thousand miles at the time, its technological improvement efficiency is much higher than Waymo.
In 2017, Cruise road test data increased by orders of magnitude, pictures from Marketwatch
Low-key stride towards the goal of fully autonomous driving
Different from Tesla’s aggressive approach of investing in the consumer market first, and then relying on sold vehicles to continuously collect test data and improve models, both Waymo and Cruise have chosen to take a more stable and stable strategic route and seek a one-step business model. , That is, Insist on first using self-produced prototype cars to overcome autonomous driving above L4 level, and then measure the production capacity to truly safely separate from human-controlled vehicles.
Cruise has been working hard all the time, and has maintained a relatively low-key attitude in the market. In addition to updating the technical blog, Kyle Vogt himself rarely accepts interviews, people just occasionally see their Chevrolet self-driving test car Bolt shuttled through the streets of San Francisco.
Cruise prototype Bolt under road test in San Francisco
Although Cruise’s goal is to achieve fully autonomous driving, it is different from Waymo’s privileged situation where Google has an unlimited budget and no return in the short term. Cruise must pay close attention to its cash reserves and actively seek commercialization possibilities.
Autonomous driving research and development itself is a money-burning machine. Cruise’s losses have increased year after year. In 2018, it lost 728 million U.S. dollars and in 2019 it lost more than 1 billion U.S. dollars. The billions of funds invested by Softbank and Honda require Cruise. It will not be credited until the commercial operation of unmanned vehicles is completed. Vogt has also repeatedly stated that the cost of self-driving cars must be quickly reduced to ensure increased consumer adoption and make Cruise’s plan financially sustainable.
In January 2019, Dan Ammann, President of GM Group, personally took charge of Cruise. This rare practice, on the one hand, shows the core role of Cruise in GM’s future technology system. On the other hand, it is also that GM wants to further strengthen Cruise’s influence and boost the process of launching autonomous driving technology into the market.
Since then, Cruise began to test the commercialization of the water. First announced the establishment of a partnership with DoorDash to deliver food and groceries to specific users in the San Francisco area, then launched a pilot project called Cruise Anywhere in San Francisco, and considered cooperating with Lyft and Uber to establish driverless cars Service Network.
In January last year, CruiseIt also released its first mass-produced model Origin for autonomous driving in San Francisco for the first time. At the press conference, Vogt personally demonstrated this fully autonomous vehicle with no driver, no pedals, and no engine.
Kyle Vogt introduces Autopilot production model Origin
In the development of Origin, we have also seen the integration attempts of old car companies with new technologies. GM is responsible for the design of the frame and electric motor, Honda focuses on the interior and interior space excavation, and Cruise is responsible for injecting sensing and autonomous driving technology into the vehicle.
Therefore, many people also believe that although Cruise is currently not as good as Waymo at the algorithm level, in the field of hardware and operations, with the strong support of GM and Honda, the overall strength of Cruise’s future cannot be understated watch for.
Now, Cruise and Microsoft are even more powerful. Microsoft’s cloud computing and edge computing platform Azure will comprehensively enhance Cruise cloud collaboration, cloud storage, and machine learning capabilities, and accelerate the pace of large-scale application of its autonomous driving solutions. With the help of Microsoft, Cruise has further widened the gap with other competitors, and also increased the confidence to fight Waymo and Tesla in the future.
After Dan Ammann became the CEO of the company, founder Kyle Vogt returned to the CTO position, focusing on technology research and development and creation, and he himself is more willing to exist as an innovator and engineer.
“Life is short. We are only awake for 16 hours a day, but many people’s time is wasted meaninglessly in the car. And I want to help people find time. This is worth our efforts. .”He said.
article from the micro-channel public number: silicon star people (ID: guixingren123) , Author: Juny , Editor: Vicky Xiao