After Baidu tie-up, BMW taps Tencent for autonomous driving in China

China is BMW’s largest market, and the German automaker knows in order to capture the country’s demanding consumers, its future models must support robust autonomous driving capability.

But to build it itself in China is hardly possible. The success of autonomous driving relies in part on high-definition mapping, a process that requires an expansive collection of geographic information. By law, foreign entities can’t host China-based data without local partnerships. Apple noticeably works with a Chinese firm to store user emails, text messages and other forms of digital footprint in the country.

That appears to be one of the catalysts for BMW’s new partnership with Tencent. The Chinese tech giant, which is best known for WeChat and runs an expanding cloud computing business, said on Friday it’s setting up a data computing and storage platform for the German premium carmaker. Reuters reported that the pair plans to launch the computing center by the end of this year in Tianjin, a port city near Beijing.

The tie-up came months after BMW’s earlier data expansion in the world’s largest passenger car market. In February, Here — a Google Maps alternative partly owned by BWM — joined forces with Chinese navigation service Navinfo which would help Here collect data locally. It’s perhaps by no coincidence that Navinfo and Tencent both bought small shares in Here three years ago.

As BMW gets more familiar with China’s road conditions, there’s no reason why it won’t apply those data to its freshly minted ride-hailing venture.

Teaming up with BMW can be a big win for Tencent, which has been placing more focus on enterprise-facing endeavors as its main gaming business copes with regulatory pressure. In the world of transportation, “Tencent is committed to assisting automotive companies in the digital transformation,” said Dowson Tong, the company’s president of Cloud and Smart Industry, in a statement.

Another relationship

BMW has previously sought after another Chinese tech leader to automate its vehicles. It has been working with Baidu, the country’s largest search engine provider with a growing list of artificial intelligence initiatives, on automated driving since 2014.

Last October, the duo ramped up their alliance after the German automaker joined Baidu’s autonomous driving open platform Apollo . The deal carried larger diplomatic significance as it came about during Chinese Premier Li Keqiang’s visit in Germany to meet with Chancellor Angela Merkel. Baidu president Zhang Yaqin said at the time the deal was meant to “accelerate the development of autonomous driving technologies that align with the Chinese market.”

BMW’s relationship with Tencent, on the other hand, has previously played out on other fronts including joint research into autonomous driving security and testing that involved Tencent’s noted Keen Security Lab.

Baidu and Tencent don’t compete directly for their core businesses, but both are making a big push into the future of mobility, whether the effort pertains to in-car entertainment or self-driving. It’s not uncommon for tech rivals in China to target the same partner. A spokeswoman for BMW told TechCrunch that “there is no overlap in the collaboration” and the German firm is “cooperating with different top-notch Chinese companies in different fields.”

Indeed, the setup with Tencent seems more comprehensive at first glance. The Chinese company is providing “IT architecture, tools and platforms supporting the entire process of [BMW’s] automated driving research and development,” according to the spokeswoman. When it comes to Baidu, she cited an example of the pair working on a self-driving safety white paper that also involved ten other partners.

That might be a roundabout way of saying that the Baidu alliance is looser. It’s worth pointing out that BMW isn’t unique to Apollo, which bills itself as the “Android for autonomous cars” and now counts more than 100 auto partners from across the world.

A large network helps generate conversations and potential leads down the road, but keeping it this way could compromise the depth of “collaboration” — a word that’s too often co-opted by publicists. As Cao Xudong, founder of Chinese autonomous driving unicorn Momenta, told TechCrunch earlier, collaboration in the auto sector “demands deep, resource-intensive collaboration, so less [fewer partnerships] is believed to be more.”

What about the other heavyweight Alibaba, which also wants to own the future of driving? The Chinese e-commerce and cloud computing company has become pally with state-owned carmaker SAIC, with which it has set up a joint venture called Banma to create autonomous driving solutions. This existing marriage means BMW will unlikely tap Alibaba for automation, an employee at a major Chinese self-driving startup suggested to me.

Snap turns to search giant Baidu to court Chinese advertisers

Two years have passed since Snap Inc first struck a deal with Baidu that authorized China’s largest search engine to be a reseller of Snapchat ads for companies in Greater China as well as Japan and South Korea, where Baidu runs a portfolio of mobile apps.

This week, the pair announced they have renewed the sales partnership without revealing how revenues are divided between the two and when the extended agreement expires.

Despite being blocked in China like most other western social media services, Snap has shown interest in China in various capacities, including a research and development center in Shenzhen for Spectacles. It’s also serving the country’s game developers, e-commerce merchants and other export-led advertisers who wish to capture the network’s 190 million daily active users around the world.

Facebook and Twitter are in the same overseas ad business in China. Facebook, with an “experience center” in Shenzhen for clients to learn how its ads work, counted China as its second-largest ad spender in 2018, according to Pivotal Research Group. Twitter also holds an annual summit in China for small and medium enterprises going global.

None of the western social giants can go it alone in China, which is why Snap chose Baidu to be its local partner to not only overcome regulatory restrictions on foreign entities but also tap the latter for language support, account management and an extensive advertiser network.

Baidu also intended to resell Facebook ads but did not manage to get a license, a former Facebook employee who wishes to remain anonymous told TechCrunch. Instead, Facebook works with Cheetah Mobile, PapayaMobile and seven other advertising representatives in China.

Through the deal, companies that purchase media through Baidu gain access to all forms of ad slots in Snap’s videos, real-time selfie effects, overlays and more. The return can be satisfying. Besides the opportunity to capture a predominantly young user base, advertisers are reaching a sticky group who, on average, opens Snapchat over 20 times and spends over 30 minutes on the app every day.

“With its young, vibrant user base, Snap’s advertising platform has been instrumental in driving growth for our game AFK Arena,” said Chris Zhang, vice president of Shanghai-based Lilith Games, in a statement.

“Our partnership with Snap Inc. provides Chinese companies new avenues to expand their businesses through Snapchat advertising,” said Sheng Hu, head of U.S. strategy and partnership at Baidu’s Global Business Unit that operates a range of overseas products such as Japanese keyboard app Simeji. “We look forward to connecting with marketing executives in China and beyond on behalf of Snap to discuss the benefits of these advertising solutions.”

The need-to-know takeaways from VidCon 2019

VidCon, the annual summit in Anaheim, CA for social media stars and their fans to meet each other drew over 75,000 attendees over last week and this past weekend. A small subset of those where entertainment and tech executives convening to share best practices and strike deals.

Of the wide range of topics discussed in the industry-only sessions and casual conversation, five trends stuck out to me as takeaways for Extra Crunch members: the prominence of TikTok, the strong presence of Chinese tech companies in general, the contemplation of deep fakes, curiosity around virtual influencers, and the widespread interest in developing consumer product startups around top content creators.

Newer platforms take center stage

GettyImages 1161447217

Photo by Jerod Harris/Getty Images

TikTok, the Chinese social video app (owned by Bytedance) that exploded onto the US market this past year, was the biggest conversation topic. Executives and talent managers were curious to see where it will go over the next year more than they were convinced that it is changing the industry in any fundamental way.

TikTok influencers were a major presence on the stages and taking selfies with fans on the conference floor. I overheard tweens saying “there are so many TikTokers here” throughout the conference. Meanwhile, TikTok’s US GM Vanessa Pappas held a session where she argued the app’s focus on building community among people who don’t already know each other (rather than being centered on your existing friendships) is a fundamental differentiator.

Kathleen Grace, CEO of production company New Form, noted that Tik Tok’s emphasis on visuals and music instead of spoken or written word makes it distinctly democratic in convening users across countries on equal footing.

Esports was also a big presence across the conference floor with teens lined up to compete at numerous simultaneous competitions. Twitch’s Mike Aragon and Jana Werner outlined Twitch’s expansion in content verticals adjacent to gaming like anime, sports, news, and “creative content’ as the first chapter in expanding the format of interactive live-streams across all verticals. They also emphasized the diversity of revenue streams Twitch enables creators to leverage: ads, tipping, monthly patronage, Twitch Prime, and Bounty Board (which connects brands and live streamers).

Andrew Ng to talk about how AI will transform business at TC Sessions: Enterprise

When it comes to applying AI to the world around us, Andrew Ng has few if any peers. We are delighted to announce that the renowned founder, investor, AI expert and Stanford professor will join us on stage at the TechCrunch Sessions: Enterprise show on Sept. 5 at the Yerba Buena Center in San Francisco. 

AI promises to transform the $500 billion enterprise world like nothing since the cloud and SaaS.  Hundreds of startups are already seizing the AI moment in areas like recruiting, marketing and communications, and customer experience. The oceans of data required to power AI are becoming dramatically more valuable, which in turn is fueling the rise of new data platforms, another big topic of the show

Last year, Ng  launched the $175 million AI Fund, backed by big names like Sequoia, NEA, Greylock, and Softbank. The fund’s goal is to develop new AI businesses in a studio model and spin them out when they are ready for prime time. The first of that fund’s cohort is Landing AI, which also launched last year and aims to “empower companies to jumpstart AI and realize practical value.” It’s a wave businesses will want to catch if Ng is anywhere near right in his conviction that AI will generate $13 trillion in GDP growth globally in the next 20 years. You heard that right. 

At TC Sessions: Enterprise, TechCrunch’s editors will ask Ng to detail how he believes AI will unfold in the enterprise world and bring big productivity gains to business. 

As the former Chief Scientist at Baidu and the founding lead of Google Brain, Ng led the AI transformation of two of the world’s leading technology companies. Dr. Ng is the Co-founder of Coursera, an online learning platform, and founder of, an AI education platform. Dr. Ng is also an Adjunct Professor at Stanford University’s Computer Science Department and holds degrees from Carnegie Mellon University, MIT and the University of California, Berkeley.

Early Bird tickets to see Andrew at TC Sessions: Enterprise are on sale for just $249 when you book here, but hurry prices go up by $100 soon! Students, grab your discounted tickets for just $75 here.

Cars-as-a-service, Alibaba and ridehailing, mental health, and the future of financial services

The future of car ownership: Cars-as-a-service

It’s Mobility Day at TechCrunch, and we’re hosting our Sessions event today in beautiful San Jose. That’s why we have a couple of related pieces on mobility at Extra Crunch.

First, our automotive editor Matt Burns is back with part two of his market map and analysis of the changing nature of how consumers are buying cars these days. Part one looked at how startups like Carvana, Shift, Vroom, and others are trying to disrupt the car dealership’s monopoly on auto sales in the United States.

Now, Burns takes a look at how startups like Fair and premium automakers like Mercedes are disrupting the very notion of owning a car in the first place. Rather than buying a car or leasing one, users with these new services are asked to subscribe to their cars, giving them the flexibility to get a car when they need it and to get rid of it when they don’t. Fair has raised $1.5 billion in venture capital, so clearly the space has caught the eye of investors.

“In simple terms,” co-founder and then CEO [of Fair] Scott Painter, told TechCrunch following its recent raise, “for every dollar in equity we unlock $10 in debt, and we borrow that cash to buy cars.”

Fair works much like a traditional lease with more options. Users can drive the vehicles as long as they’re paying for them and can switch to a different one whenever. This is different from a traditional lease where the buyer is often locked into the vehicle for two to four years. The model makes Fair an excellent option for Uber and Lyft drivers, and in the last year, Uber sold fair its $400 million leasing business to accelerate this offering.

Meituan, Alibaba, and the new landscape of ride-hailing in China

Meanwhile, on the other side of the world, our China tech reporter Rita Liao takes a deeper look at the quickly changing tides of the ride-hailing industry in China. It’s a fight between intermediation, disintermediation, and who ultimately owns the ride-hailing consumer. As transit in China and the rest of the world increasingly becomes multi-modal, who owns the gateway to figuring out the best method and paying for it is increasingly in the driver’s seat:

Meituan, Alibaba, and the new landscape of ride-hailing in China

Instead of switching between apps to secure a ride during rush hour, people in China can now hail from different companies using a single app. Some of the country’s largest internet companies — including ride-hailing giant Didi itself — are placing bets on this type of aggregation service.

The nascent model is reminiscent of a feature Google Maps added in early 2017 allowing users to hail Uber, Lyft, Gett and Hailo straight from its navigation app. A few months later, AutoNavi, a maps app owned by Alibaba, debuted a similar feature in China. Other big names like Baidu, Hellobike, Meituan and Didi subsequently joined forces with third-party ride-booking services rather than building their own.

The trend underscores changes in China’s massive ride-hailing industry of 330 million users (in Chinese). The government is tightening rules around vehicle and driver accreditation, leading to a widescale driver shortage. Meanwhile, established carmakers including BMW and state-owned Shouqi are entering the fray, offering premium rides with better-trained fleet drivers, but they face an uphill battle with Didi, which gobbled up Uber China in 2016.

By corraling various ride-booking services, an aggregator can shorten wait time for users. For new ride-hailing players, riding on a billion-user platform like Meituan opens up wider user acquisition channels.

These ride-hailing marketplaces let users request rides from any number of third-party services available. At the end of the trip, users pay directly through the aggregator, which normally takes a commission of about 10%, although none of the players have disclosed how revenue is exactly divided with their mobility partners.

In comparison, a ride-hailing operator such as Didi charges about 20% from each trip since they take care of driver management, customer support and other dirty work which, to a great extent, helps build the moat around their business.

Here’s a look at who the aggregators are.

Intel and Baidu partner on Nirvana Neural Network AI training processor

At Baidu’s Create conference for AI developers in Beijing today, the company and Intel announced a new partnership to work together on Intel’s new Nervana Neural Network Processor for training. As its name very clearly states, this forthcoming chip (NNP-T for short) is a processor built specifically for the task of training neural networks for the purposes of performing deep learning at scale.

Baidu and Intel’s collaboration on the NNP-T involves working together on both the hardware and software side of this custom accelerator to ensure that its optimized for use with Baidu’s PaddlePaddle deep learning framework, which will complement existing work that Intel has already done to ensure that PaddlePaddle is set up to perform best on its existing Intel Xeon Scalable processors. The NNP-T optimization will specifically focus on applications of PaddlePaddle that focus on distributed training of neural networks, to complete other types of AI applications.

Intel’s Nervana Neural Network Processor lineup, named after ‘Nervana,’ the company it acquired in 2016, is developed by the Intel AI group led by former Nervana CEO Naveen Rao. The NNP-T is tailor-made for training AI (ingesting data sets and learning how to do the job its supposed to do), while the NNP-I (announced at CES this year) is designed specifically for inference (taking the results of the learning process and putting into actions, or actually doing the job it’s supposed to do).

The NNP made its debut in 2017, and the first-generation chip is currently being used as a software development prototype and demo hardware for partners, while the new so-called ‘Spring Crest’ generation are targeting production availability this year.

How safe are robotaxis? BMW, Intel, Aptiv (and 8 others) just laid out a safety blueprint

Self-driving vehicles are often trumpeted as the answer to the million of injuries and deaths that occur each year on the world’s roadways.

Developers and proponents of the technology argue that an automated system that can see, hear, react and make better decisions than humans will reduce the number of crashes that occur every year. And that’s a considerable promise. Some 1.35 million people died in vehicle crashes in 2016, according to the most recent statistics from the World Health Organization.

Still, there is no way for companies to guarantee that every self-driving vehicle put on the world’s roads can guarantee the same robust level of safety and security.

Eleven companies have formed a consortium to change that.

Tier 1 suppliers Aptiv and Continental, automakers Audi, BMW, Daimler, Fiat Chrysler and Volkswagen as well as chipmakers Intel and Infineon, mapping consortium HERE and China’s internet search giant Baidu released Tuesday a 157-page white paper that outlines how to build, test and operate a safe automated vehicle. (the white paper is embedded below)

The idea behind the “Safety First for Automated Driving,” group (SaFAD for short) is to create a blueprint of sorts that lays out 12 principles for designing — and later testing and validating — safe automated vehicles, Karl Iagnemma, president of Aptiv Mobility told TechCrunch.

This isn’t meant to be a static effort. One technological breakthrough could render portions of the white paper moot. The intent was to create a “living” document that grows and adapts along with technology and the industry, Iagnemma said.

Nor does the document pick technological winners or losers. Those who sift through the white paper won’t find endorsements for certain sensors, for example. Instead, the group is pushing for common ground on safety.

The papers lays down certain “safety by design” rules that engineers should be thinking about and following from the beginning. This first section of the document lays out 12 guiding design principles such as proper cybersecurity, protections if the system degrades or fails, operational design domain and data recording. A number of the principles deal with how a vehicle operator and the automated driving system interact with each other, including ensuring that if a handover does need to occur (this would be in vehicles that could switch between manual and automated controls) that it is properly and explicitly communicated to the human.

The principles also delve into the behavior of the automated vehicle to ensure that its predictable for the user and easy-to-understand for surrounding road users like bicyclists, pedestrians and other drivers.

The architects of this framework even weigh in on the safety of the interior of these automated vehicles. It’s seemingly small, yet important guideline. So many of the futuristic concept shown at tech and auto shows depict lounge-like interiors where people can face each other. The guidelines simply state that occupants should be protected even when there are new uses for the interior.

The second half of the document covers the validation and verification of these systems. In other words, how do you make sure that the automated vehicle you designed and built actually works? The consortium believes proper validation and verification should include testing on tracks, open roads and in simulation.

One more interesting note found within the pages of the white paper is an appendix that deals with machine learning and more specifically deep neural networks, namely how to ensure they’re developed, deployed and properly validated.

All of this, of course, sounds like common sense. And yet, companies have been working on automated driving systems without a single standard to guide them.

Many companies might be following a similar rules laid out in the framework released Tuesday, but it’s impossible to know exactly what they’re all up to, especially when considering this on a global scale. The goal of the framework’s architects is to get as many companies as possible to sign on until they hit a critical mass within the industry.

Sure, companies have signed onto consortiums like this before to push for safe automated vehicles or promote public education campaigns. It’s worth noting that this effort was apparently driven by engineers within these companies, and not a marketing team or board of directors.

It should be noted that there is an existing international standard called ISO 26262 that covers the functional safety of electrical and electronic systems in production vehicles. And another standard ISO 21448 or “Safety of the Intended Functionality” is in the works to handle advanced driver assistance systems and autonomous vehicles. The SaFAD consortium as well as another group the European Association of Automotive Suppliers, or CLEPA), which is being led by Nvidia, are involved in the development of ISO 21448.

The hope is that out of these various efforts, a single, clear standard rises to the top that can be validated and verified. Without it, companies may have trouble building trust with a wary and fickle public.

California has let two Chinese startups offer robotaxis to the public

China’s driverless cars are coming for passengers in the United States. AutoX and just became the first Chinese companies allowed to offer fully self-driving cars in the state of California, according to notices posted on the website of the California Public Utilities Commission this week.

Started in 2016 by Princeton University professor Jianxiong Xiao, called “Professor X” by his students, AutoX is now one of China’s most well-funded autonomous driving startups alongside, which was co-founded in 2016 by two former executives at Baidu’s self-driving department.

AutoX said in January that it was in talks with investors to raise a lofty $100 million. had banked at least $214 million in funding as of April.

While more than 62 companies hold the permits to test autonomous vehicles in California, very few are actually allowed to transport people in those cars. Zoox passed a new milestone when it received the first green light to provide robotaxi services in the state six months ago. Now AutoX and have joined the exclusive club, bringing the number of participants in the pilot program to three.

autonomous driving

Screenshot from the California Public Utilities Commission website

There are a few catches though. The type of permission granted to the three companies is for the “Drivered AV Passenger Service,” which forbids companies to charge passengers for test rides and requires a safety driver behind the wheel. No entity has so far been permitted to run real driverless passenger service in California, a sign that regulators aren’t quite ready to let tech companies transport the public without human oversight.

AutoX, which is already using self-driving vehicles to deliver groceries in San Jose, is getting a headstart by introducing California’s first robotaxi service. People living in north San Jose or Santa Clara can now apply to join its early rider program and give feedback, says an instruction on its website. A spokesperson for told TechCrunch that the company also began offering driverless passenger services as soon as it received the permit.

Alphabet’s Waymo launched a passenger service in Phoenix last December. Like California, Arizona demands a trained test driver to assist with operations if needed. While Waymo is allowed to charge passengers, it can only ferry a vetted group of people, so the program isn’t available to everyone.

These confinements seem sensible given legal and ethical concerns raised by critics. Last year, Uber’s self-driving test vehicle struck and killed a woman in Temple, prompting the transportation giant to suspend its test drives. The incident has become a cautionary tale for startups in the field. Take Momenta, the first Chinese autonomous driving startup to pass $1 billion in valuation. The CEO requires all executives to ride a minimum number of autonomous miles themselves, so the management would put passenger safety first.

Chinese startups covet the California license for a number of reasons. First, self-driving cars are by nature data-hungry. There are only a small handful of cities worldwide which allow robotaxis on public roads, so it always helps to collect more mileages whenever permissible.

Many of these Chinese companies have also set up research and development centers in California to tap the region’s tech talent., for example, deploys R&D staff and offices across Silicon Valley, Beijing and Guangzhou. AutoX opened an R&D center in Shenzhen earlier this year but still keeps development teams in San Jose.

China’s housing unicorn Danke appoints ex-Baidu exec as new COO

A few months after nabbing a handsome $500 million funding round, China’s shared housing startup Danke Apartment got a talent boost.

On Monday, Danke announced the appointment of Gu Guoliang as its new chief operating officer to ramp up the company’s offline operational crew. Gu, whose nickname is Michael, stepped down from Baidu after five years as one of the key figures in search, historically the company’s biggest revenue-generating division. He’s known to have managed several tens of thousands of marketing staff and helped generate sales of close to 100 billion yuan ($14.44 billion) for Baidu annually.

Gu’s arrival followed a period of explosive expansion at Danke, which is now managing almost 500,000 units of rooms across 10 Chinese cities after founding four years ago. The startup takes the co-living approach akin to that of WeWork’s Welive and rents out fully furnished apartments targeted at young professionals who can’t afford a full suite. Backed by Tiger Global and Alibaba’s financial affiliate Ant Financial, Danke’s valuation crossed $2 billion in its funding round in February.

Gu is one of the former Baidu executives who resigned during a recent top-level exodus (report in Chinese) that involved at least five leaders, including the search division boss Xiang Hailong, to whom Gu reported. There were speculations that Xiang’s exit might have triggered his lieutenants to leave, though TechCrunch has learned from a person close to Gu that he had left “one to two weeks” prior to Xiang’s departure.

For Gu, joining Danke would almost feel like returning home. “We welcome our comrade and good friend Michael,” said Danke chief executive Gao Jing, who previously worked alongside Gu at Nuomi, the local services startup that was sold to Baidu for $3.2 billion and became integral to the internet giant’s online-to-offline business. Derek Shen, an investor and current chairman of Danke, co-founded Nuomi in 2010 before heading up LinkedIn China between 2014 and 2017. Several other core members of Danke have also hailed from Nuomi.

Danke is confident that Gu’s addition will be a boon to its operational capacity. “Gu has abundant experience in operational management, sharp business insights, outstanding leadership, and a deep understanding of the internet sector and user needs,” said Gao. “Under his direction, Danke will enter a new phase of refined operation.”

By that, Gao means Gu will be tasked with rolling out more targeted marketing, more efficient housing renovation, more precise acquisition of apartment space, among other quality-control measures to drive sustainable growth at the company.