Beyond trendy names like Tesla and Alphabet chasing self-driving cars, a host of auto brands and other tech heavyweights are also investing in autonomous R&D.
Private companies working in auto tech are attracting record levels of deals and funding, with autonomous driving startups leading the charge.
Along with early-stage startups, VCs, and other investors, large corporations are also angling to get a slice of the self-driving pie.
Download THE PDF
Using CB Insights’ investment, acquisition, and partnership data, we identified over 40 companies developing road-going self-driving vehicles. They are a diverse group of players, ranging from automotive industry stalwarts to leading technology brands and telecommunications companies.
This list is organized alphabetically and focuses on larger corporate players in the space (as opposed to earlier-stage startups). Companies working on industrial fully autonomous vehicles were not included in this analysis.
A few of the companies or brands listed below belong to the same parent organization but are detailed separately if they are operating distinct autonomous development programs. Some companies are grouped together by key partnerships or alliances. Given the complex web of relationships between these players, other collaborations are also noted in each profile.
This is not intended to be an exhaustive list of corporations working on autonomous vehicle technology.
This brief was originally published on 9/25/2015 and featured 25 select corporations. It was updated and expanded on 5/17/2017, 9/4/2018, 8/28/2019, 3/4/20, and 12/16/20.
Amazon experiments with autonomous package delivery
- Filed patent in 2015 for autonomous lane-switching technology
- Invested in autonomous tech startup operated by Google and Tesla veterans
- Autonomous delivery robot Amazon Scout currently making the rounds in several US cities
- Working on a multi-function autonomous vehicle with Toyota
Over the last decade, Amazon has spent billions of dollars working on finding ever-better solutions to the last-mile problem in delivery. It’s built its own fleet of cargo jets, explored delivery by drone in the form of “Prime Air,” and more. More recently, an increasing percentage of that investment has been directed toward autonomous vehicle technology.
In September 2019, Amazon said that it planned to put 100,000 electric delivery vehicles on the road by 2030, with deliveries starting as soon as 2021. The US technology giant plans to buy these vehicles from Rivian, an electric automaker in which Amazon has invested $440M. Amazon revealed a prototype of these delivery vehicles in January 2020.
Amazon’s investment in Rivian came on the heels of its deal in February 2019 with Aurora Innovation, an autonomous tech startup run by former Google and Tesla executives that pioneered autonomous driving teams at both giants.
An Amazon company spokesperson told Wired about the investment:
“Autonomous technology has the potential to help make the jobs of our employees and partners safer and more productive, whether it’s in a fulfillment center or on the road, and we’re excited about the possibilities.”
The Rivian and Aurora investments aren’t the only autonomous technology play that Amazon is pursuing. In January 2019, the company introduced the Amazon Scout, a 6-wheeled electric-powered delivery robot. These robots are currently making deliveries in a Washington neighborhood and in Southern California’s Irvine area during daylight hours, Monday through Friday, under the supervision of a human associate. In July 2020, the project expanded its geographical coverage. Amazon’s delivery robots will now serve select customers in Franklin, Tennessee and Atlanta, Georgia.
Additionally, in June 2020, Amazon announced plans to acquire Zoox, a startup developing autonomous driving tech geared toward ride-hailing customers. This $1.2B deal provides the e-commerce giant with access to software, AI, and a full-stack self-driving solution. Backed by Amazon, Zoox is now better equipped to take on its primary rival, Waymo, an Alphabet subsidiary. In December 2020, Zoox unveiled its fully autonomous, electric robotaxi, which can seat 4 passengers and reach speeds of up to 75 mph. The vehicle is currently in testing.
For Amazon, these developments are years in the making.
The company initially announced that it would be getting involved in autonomous vehicles at CES 2018 through a partnership with Toyota. The demo vehicle, known as the e-Palette, was designed as a multi-function, autonomous minivan to move goods, people, or even a mobile office. The plan was to debut up to 20 of its e-Palette vehicles at the 2020 Summer Olympics and Paralympic Games in Tokyo to transport athletes. But the Covid-19 pandemic forced Japanese organizers to postpone these sports events until July 2021.
In 2015, Amazon explored a trial with DHL and Audi that involved delivering customers’ parcels to the trunks of their automobiles.
In the same year, Amazon filed for a patent for a system that helps autonomous cars navigate roadways, especially complex, reversible lanes, and even pick which lane to use depending on current traffic estimates — an early indicator of the company’s ambitions in this space and AVs’ importance to Amazon when it comes to lowering delivery costs. The company was granted the patent in early 2017.
In April 2017, the Wall Street Journal reported that Amazon had built a team more than a year prior devoted to focusing on driverless vehicle technology.
Read more: The Verge, CNBC, Forbes, Business Insider, TechCrunch
Apple is investing in research, testing, and acquisitions for autonomous driving
- Building employee transportation network; currently has 70 self-driving vehicles on the road in California
- Hired former Waymo and NASA engineer in June 2018 to head Project Titan
- Acquired Drive.ai in June 2019
- Secured a number of autonomous driving patents
In January 2019, Apple cut more than 200 employees from its self-driving car initiative, Project Titan, in what was internally described as a “restructuring.” Five months later, Apple confirmed that it had acquired Drive.ai, a self-driving startup that is backed by more than $77M in funding and was valued at $200M in 2017. The move is reported to be an “acqui-hire,” with Apple’s interests lying more with Drive.ai’s talent than its proprietary technology.
For Apple, the move appears to be an effort to get its own autonomous vehicle efforts back on track. Project Titan has suffered setbacks since as early 2016, which marked the departure of project head Steve Zadesky and a rumored hiring freeze, as well as strategic uncertainty about the vision of the project.
In July 2016, Apple selected its legendary hardware executive Bob Mansfield to lead its effort, in addition to hiring Dan Dodge, the founder and former CEO of QNX. The hires indicated a shift in strategy, with Project Titan reportedly deciding to prioritize the development of an autonomous driving system, while deprioritizing development of an electric vehicle.
In April 2017, new details seemed to confirm this pivot: Apple documents revealed the company was building an “automated system,” and the company hired robotics experts from NASA to boost its driverless efforts.
Apple spent 2018 building out the beginnings of its self-driving car fleet, with 70 vehicles officially on the road and registered with the California DMV as of September of that year. This reportedly made Apple the owner of the third-largest autonomous test vehicle fleet in the state, behind GM Cruise and Waymo. Apple’s test fleet logged a total of 80,739 miles of autonomous driving time from April 2017 through November 2018. In the following 12 months, however, the program slowed, logging only 7,544 miles.
In May 2018, Apple was reportedly working to provide autonomous cars for employee transit between corporate facilities. The Volkswagen T6 transporter vans slated to be used will feature a human driver present in case of any issues with the self-driving technology in the car.
In June 2018, Apple brought on Jaime Waydo, a senior engineer with experience at both Waymo and NASA’s Jet Propulsion Laboratory, to work on Project Titan.
In late 2018, Apple poached a designer and vehicle engineer from Tesla, a move that suggested that it may be reevaluating Project Titan’s move away from building an autonomous vehicle. The company has also built prototypes of an electric van, Apple Insider reported in February 2019.
In December 2020, Apple transitioned Project Titan to its AI and machine learning group led by John Giannadrea, according to Bloomberg.
The Cupertino-based giant secured a number of patents with the US Patent and Trademark Office in September 2020. One outlines a solution that recognizes when optical sensors fall out of alignment due to bumps in the road or vibrations and adjusts them automatically. Another patent describes a mechanism for steering during autonomous driving.
CB Insights clients can view Apple’s vehicle-related patents here.
Read more: Gizmodo, The Guardian, Reuters, Bloomberg, The Guardian, Wall Street Journal
Aptiv logs 100,000+ self-driving rides and sets sights on robotaxi fleets
- Autonomous driving operations underway in Boston, Las Vegas, Pittsburgh, and Singapore
- Completed 100,000+ self-driving taxi rides in Las Vegas in collaboration with Lyft
In February 2020, self-driving software company Aptiv and ridesharing company Lyft reached a major milestone: the companies crossed 100,000 rides of its autonomous passenger service in Las Vegas. The company now services over 3,400 spots in Las Vegas, including popular locations such as the Los Angeles Convention Center and McCarren International Airport.
The program features a fleet of BMW 540s, retrofitted with Aptiv’s autonomous tech but with a safety driver present in the car. Reception to the program has been positive: Lyft reports that 98% of the passengers who took the rides gave the service a 5-star rating.
The Lyft partnership is just one of several programs that Aptiv has underway with its self-driving software. The company currently has autonomous driving operations in 4 markets: Boston, Las Vegas, Pittsburgh, and Singapore. In spring 2019, the company announced it was adding a fifth city, Shanghai.
The expansion into China is a major strategic move for Aptiv. China is expected to account for more than two-thirds of autonomous miles driven worldwide by 2040, according to McKinsey.
Aptiv’s President of Autonomous Mobility Karl Iagnemma said:
“The long-term opportunity in China is off the charts… For Aptiv, it’s always been a question of not ‘if’ but ‘when’ we’re going to enter the Chinese market.”
These initiatives have translated into strong financial results for Aptiv. Over the course of 2019, the company’s stock posted gains of more than 50%. Stock performance was impressive in 2020 as well, with per share price reaching around $123 in December 2020.
Another major Aptiv initiative is Motional, its autonomous driving venture with Hyundai. The venture was officially unveiled in August 2020. Motional aims to build and commercialize SAE Level 4 vehicles (autonomous vehicles that can perform all driving functions), and make this technology available for robotaxi fleets in 2022. To make these plans a reality, Motional plans to increase its staff count to over 1,000 people in 2020. Karl Iagnemma, Motional’s president, says that his team “will have a car that can drive itself in defined areas ready for external testing by 2022.” In late November, the venture received approval from the state of Nevada to test its vehicles on public roads without safety drivers.
The last few years have witnessed a number of transformations at Aptiv. In May 2017, the company, then known as Delphi Automotive, spun off of its powertrain segment and rebranded itself as a new company: Delphi Technologies. A few months later, that company was renamed Aptiv. The rebrand was intended to focus the company’s software, electrical components, and other work on autonomous vehicles.
In October 2017, Delphi acquired autonomous software development shop nuTonomy, Inc. for $400M upfront. The goal of the acquisition was to get more than 60 Aptiv-branded autonomous cars onto public roads by the end of the year. The acquired startup had a strong track record: in 2016, nuTonomy integrated self-driving car technology into a ride-hailing service for the very first time in Singapore. It then also became the first to test such technology on public streets in Boston, Massachusetts. In April 2018 (after the acquisition), nuTonomy was officially named a finalist for Fast Company’s 2018 World Changing Ideas Awards.
The company has also partnered with Quanergy Systems to develop and deploy solid-state lidar, which could dramatically lower the cost of these systems.
Read more: Reuters, CNET, Ars Technica, Delphi, AP
Audi unveils its autonomous A8, and its self-driving tech system merged with Argo AI
- First auto company to deploy hands-free driving
- Flagship self-driving A8 model approved for street driving in Europe
- Former Tesla Autopilot manager hired as CTO of self-driving tech subsidiary
- Audi’s Autonomous Intelligent Driving (AID) unit merged with Argo AI
Audi has made big promises about its plans for autonomous and electric vehicles. The German brand has said it plans to spend close to $16B on self-driving and sustainable tech by 2023.
Those efforts were primarily being undertaken by Autonomous Intelligent Driving (AID), Audi’s self-driving technology outfit, first launched in 2017. The operation, which is based in Munich and boasts more than 200 employees, had 12 autonomous vehicles on public roads in its home city in 2018.
Audi chief technology officer Alexandre Haag has compared AID to GM’s self-driving unit, Cruise, and Ford’s unit, Argo. But Haag is pragmatic about the realities of getting autonomous fleets deployed, given how much rests on the dependability — and safety — of the technology. He said:
“Getting to 90 percent [in perception] is fairly easy. Getting to 95 percent starts to get interesting. And then you still need to go way beyond that. Nine point nine nine nine nine… Adding each nine is ten times harder. When you’re at 95 percent, you’ve just scratched the surface.”
Audi formed new technology partnerships in an effort to expedite its autonomous timelines. In December 2018, the company announced that AID would partner with Luminar, a Silicon Valley manufacturer of lidar sensors and perception software that also works with Volvo and Toyota. But in June 2020, AID was acquired by the US-based startup Argo AI — jointly owned by Ford and Audi’s parent Volkswagen — and became its European headquarters and fifth engineering center.
Audi is also part of the German consortium — which includes Daimler and BMW — that bought Nokia’s HERE precision mapping assets for $3.1B. HERE has made strides in designing an open specification for vehicle sensor data collected and transmitted to the cloud by connected vehicles.
Audi has revealed several autonomous vehicle prototypes derived from its A7 and RS 7 models, including consumer-oriented test vehicles. In July 2016, news broke that Audi was joining many automotive peers by setting up its own advanced subsidiary, SDS Company, to focus on self-driving tech. In April 2017, Audi hired former Tesla Autopilot program manager Alexandre Haag as the unit’s CTO. In June 2020, he was appointed the head of technology for Argo AI’s Munich office.
Audi plans to commercialize its technology in its next-generation A8 flagship, although the vehicle’s SAE Level 3 automation will have limited availability, pending regulatory approval. The luxury brand operates under the umbrella of the Volkswagen Group, so developments within the division could have broader implications going forward, as evident in the case of AID and Argo AI.
While a fully autonomous fleet may still be a ways away for Audi, the company has been among the most aggressive at introducing semi-autonomous options to the market.
In July 2017, Audi unveiled its new flagship A8, which has a controversial autonomous driving feature that lets drivers fully take their hands off the wheel while the car drives up to 37 mph. At the time, it was the first vehicle in production that could actually allow its users to “drive” hands-free.
The 2019 A8 model includes an enhanced suite of semi-autonomous features, including adaptive cruise control and collision mitigation capabilities, as well as improved object recognition.
The Audi A8 with the full complement of self-driving features has only been initially approved for release and made street-legal in Europe. But the company seems to be taking a step back: its A8 sedan won’t be equipped with the Level 3 partial automation system, called Traffic Jam Pilot, due to safety concerns and the lack of a legal framework.
Read more: TechCrunch, Wired, Digital Trends, USA Today
Autoliv reimagines the future of airbags in autonomous vehicles
- Developing an advanced airbag that adapts to the interior of autonomous vehicles
- Teamed up with Volvo to develop autonomous car software
As AVs are gradually deployed in the decades ahead, they will share the road with conventional vehicles and fallible human drivers. The promise of safer roads enabled by self-driving tech will take time to materialize. Carmakers are developing various safety solutions to account for inevitable crashes.
Autoliv, the world’s largest automotive safety supplier, is at the forefront of these efforts. The company is trying to reimagine the future of airbags and prevent human errors from turning into catastrophes. For example, drivers of autonomous vehicles may be able to swivel in their chairs, lower seats into beds, or turn around to talk to other passengers. In the event of an accident, typical seatbelts and airbags would be of little use.
Autoliv is developing an airbag that resembles a protective cocoon and is integrated into the seat frame. It would protect drivers from loose items or unbuckled backseat passengers. Autoliv showcased a prototype in January 2019 at the AutoMobili-D exhibition in Detroit.
The company was also working with Volvo on creating self-driving software. In 2017, the duo established a new joint partnership called Zenuity that uses Nvidia’s AI technologies for software development. Three years later, Zenuity was split in two. Volvo took over the autonomous driving division, while Autoliv’s spinoff Veoneer will continue work on advanced driver assistance software. Both parent companies will retain access to Zenuity-owned technologies.
We’ve written more about the future of the in-vehicle experience and safety technologies here.
Download THE PDF
Baidu’s Apollo platform logs 1M+ autonomous miles driven and 150+ partners
- Built “Android of autonomous driving” to support development of self-driving tech
- 3M autonomous kilometers driven in urban environments
- Opened AI research lab in Silicon Valley
- Launched a robotaxi service in Beijing
The Chinese autonomous vehicle market represents a huge opportunity — $500B by 2030, according to McKinsey. Companies around the world are jockeying to capture that market, including Chinese tech giants Tencent and Alibaba. But only one contender can claim nearly 2M miles of autonomous driving: Chinese search company Baidu.
The main hub of Baidu‘s automation efforts is Apollo, its open-source autonomous driving platform, which originally launched in 2017. Baidu’s COO has called Apollo the “Android of the autonomous driving industry.” More than 150 partners worldwide use the service, including automakers Chevy, Ford, Honda, Toyota, and Volkswagen, as well as technology company Intel, according to Baidu.
Baidu first began testing its open-source Apollo software system on the open road in 2017. In 2018, the company received approval from the Chinese government to start testing Apollo on 33 different roads spanning 65 miles around Beijing. The Beijing Municipal Commission of Transport also granted Baidu its first batch of T4 test permits — the highest level of autonomous vehicle clearance currently available in China. The T4 permits let Baidu operate its vehicles in challenging urban environments, such as tunnels and school zones. In January 2019, Baidu unveiled Apollo Enterprise, a new line that will focus primarily on use cases including highway driving, valet parking, autonomous mini-buses, and mapping technology. These efforts have yielded impressive results.
By December 2019, Baidu had 300 autonomous vehicles on the road and had logged over 3M kilometers (roughly 1.8M miles) of urban driving. The same month, Baidu received licenses to test its fleet of autonomous robot taxis in certain areas of Beijing. Nine months later, the company launched a robotaxi service in China’s capital. A fleet of 40 cars serves almost 100 pickup and drop-off stations, and rides can be summoned via Baidu’s e-hailing platform.
In August 2020, Baidu announced that its Apollo Computing Unit (ACU) was ready for use. The autonomous driving computer is capable of handling huge amounts of radar and camera data. It will power a valet parking service called Apollo Valet Parking, similar to Tesla’s Summon feature. ACU will initially be deployed in electric vehicles built by the China-based carmaker WM Motor.
Baidu has also entered a partnership with Zhejiang Geely Holding Group to equip Geely cars with self-driving technology provided by Baidu, including Apollo’s DuerOS. Tesla partnered with Baidu in January 2020 to use its mapping data for its vehicles in China.
In April 2014, Baidu partnered with BMW to develop a semi-autonomous prototype. The partners tested their technologies on highways in China but parted ways in November 2016 following disagreements about strategy.
The search giant has also opened two Silicon Valley artificial intelligence research labs, although Andrew Ng (its chief AI scientist) departed the company in March 2017. Baidu also intends to spin off its self-driving unit once it matures (similar to Alphabet with Waymo).
For years, Apollo has been used by chipmakers and small startups to kickstart work in self-driving cars, but big corporates — including Nvidia, Bosch, Daimler, and Ford — have also found it useful.
Read more: Venture Beat, Guardian, BBC, SCMP
BMW-Intel-Mobileye alliance for autonomous car projects
- Showed off autonomous car concept at CES 2016
- Plans to deploy self-driving car on the road by 2021
- Opened second autonomous driving campus in Munich in 2018
BMW has begun aggressively pushing its autonomous strategy since showing off an autonomous i8 concept at CES 2016. In July 2019, the company announced a partnership with another German automaker, Daimler. The companies agreed to commit 1,200 technicians to the task of developing new autonomous systems with the goal of getting them on the road by 2024. But this collaboration ended less than a year later. The complexities and expenses of building a joint tech platform proved to be larger than the duo expected.
Munich-based BMW has been pursuing many other partnerships, too. In 2016, the company announced an alliance with Intel and Mobileye, which Chrysler and Magna also joined in 2017. The coalition plans to create an open standards-based platform for bringing self-driving cars to market, aiming to put its first vehicle, the BMW iNEXT, on the road by 2021, through recent reports suggest Level 3 partial automation (or above) will not be available at market launch.
Intel has been keen to push into the sector, having been beaten to the punch by companies like NXP and Nvidia to supply automotive silicon and autonomous processing power. In late 2016, it created a new Autonomous Driving Group (ADG) and committed $250M to auto tech investments through its Intel Capital arm. In March 2017, Intel further ramped up its autonomous focus when it announced its $15.3B acquisition of Mobileye.
Mobileye itself has a number of other partnerships through its Road Experience Management (REM) mapping platform, including Nissan, VW, and BMW. BMW is also part of the group that bought Nokia’s HERE mapping assets for $3.1B. In 2017, Intel took a 15% stake of HERE as well.
In 2018, BMW opened an autonomous driving campus near Munich, Germany, to work on self-driving pilot projects, making it BMW’s second workplace dedicated to autonomous technology alongside its office in Mountain View, California.
But the company is incorporating software from other smaller startups as well. In September 2020, it inked an agreement with Tactile Mobility, an Israel-based company whose technology detects road conditions by analyzing wheel speed, gear position, and other non-visual sensors.
Read more: CNBC, BMW Group, WSJ, Reuters
Bosch tests autonomous taxi service and automated valet parking
- Began trial for a self-driving taxi service in California
- Invested $1.1B in a facility for producing semiconductors that will be used in self-driving, smart home, and smart city technologies
- Developed world’s first fully autonomous parking function approved for everyday use with Daimler
- Internally dedicated 2K+ engineers to working on driver-assist technology
In December 2019, Bosch and partner Daimler started a trial for a self-driving taxi service in San Jose, California, a first for both the German companies. The companies said that a select group of users will be able to call for a Mercedes-Benz S-class vehicle through an app, to ply on a fixed route in the city. The passengers will be accompanied by a human driver to monitor the vehicle.
The trial comes as part of a collaboration between Bosch and Daimler that stretches back to 2015. The companies have also announced plans for urban road-friendly vehicles to be brought to market by the early 2020s.
In July 2019, Bosch and partner Daimler received approval from German regulators to operate their autonomous parking feature without having a human safety driver behind the wheel. The approval marks the world’s first fully automated parking function to be approved for everyday use.
Bosch also teamed up with Ford on a similar project. In August 2020, the companies demonstrated an automated valet parking solution in downtown Detroit. A Ford Escape test vehicle “talked” with Bosch sensors to locate and park in an empty spot. These automated parking solutions can accommodate up to 20% more vehicles and deliver other services, such as refueling and charging.
Bosch, one of the world’s largest automotive suppliers, reportedly has more than 2,000 engineers for driver-assistance systems. The company is also in the process of constructing a $1.1B facility in Dresden that will produce semiconductors for use in autonomous vehicles, as well as smart home and smart city technologies. Production at this facility is set to begin in 2021.
In October 2019, Bosch became a part of a consortium working towards making self-driving technology commercially viable. Some other members of the Autonomous Vehicle Computing Consortium include Toyota, GM, Continental, Nvidia, and NXP.
Daimler-owned Mercedes-Benz has also been taking other steps toward self-driving cars. Like other luxury brands, Mercedes-Benz has begun deploying semi-automated advanced driver assistance systems in many of its newer models.
An ad characterizing Mercedes’ 2017 E-Class as “self-driving” landed the company in hot water, and it pulled the ad in the face of fierce criticism from consumer advocates. Like Tesla’s Autopilot, the E-Class was capable of Level 2 (partial) automation.
In April 2017, Bosch and Mercedes-Benz joined forces to develop Level 4 (high automation) and 5 (full automation) vehicles, with Mercedes-Benz having 2 years of exclusivity to the co-developed system before it can be offered to competing automakers.
Read more: TechCrunch
Cisco works on building the data layer of the self-driving car movement
- Started building autonomous driving infrastructure with Michigan DOT in 2017
- At CES 2018, announced a project to build technology bringing gigabit-speed connectivity to smart cars
- Testing OpenRoam platform for uploading and downloading data produced by autonomous vehicles
Over the last few years, Cisco has begun exploring the potential in building out the data collection and analysis layer of the autonomous driving and smart car industries.
In October 2017, Cisco began working with the State of Michigan’s Department of Transportation on a self-driving car “adjacent” initiative. The purpose of the project, also known as Cisco Connected Roadways, was to find ways to better connect individual automobiles on the road with the infrastructure around them — including roads, parking meters, and street lights.
At CES 2018, Cisco announced that, through a partnership with Hyundai, it would focus on bringing gigabit-speed Ethernet connectivity to smart cars — enabling both faster-than-ever over-the-air updates and setting the foundation for better self-driving technology.
Cisco says its Ethernet technology should allow automakers to save $35B over the next 4 years by eliminating the need for many routine dealership trips. It could also boost the development of autonomous technology, given how much information self-driving cars need to be able to process and send for analysis every second they are monitoring the road.
The networking giant is also testing its OpenRoam platform in close partnership with UK-based startup Oxbotica. The platform acts as an infrastructure that autonomous vehicles fitted with Oxbotica software could use to upload their data to the cloud. The rapid upload and download will be done through Wi-Fi hotspots at parking garages, gas stations, and more.
Read more: Business Insider, CNBC, ZDNet, Network World
Continental AG takes an incremental approach to developing autonomous driving technologies
- Focus on driver-assist technologies and smart car infrastructure
- Opened Silicon Valley R&D lab in 2017
- In 2018, announced partnership with Nvidia to build self-driving vehicle systems
- Plans to invest €100M ($122M) in building a Texas-based radar sensor factory
German auto supplier Continental AG has steadily been operating its own autonomous vehicle program. The company has been taking a gradual approach to self-driving cars, only committing to a vague “2020s” time frame for its products and preferring to gradually roll out driver-assist technologies such as its “Cruising Chauffeur.”
In February 2020, Continental said it would invest €100M ($122M) to build a plant for radar sensors in Texas. The investment comes as part of the company’s plan to invest “high triple-digit million euros” in the areas of assisted and automated driving over the next 5 years.
At the company’s annual Tech Show in July 2019, Continental self-driving head Andree Hohm highlighted a number of barriers that made faster implementation of autonomous technologies untenable. These include continued technological limitations, consumer skepticism, and the variation in how technologies and regulations are taking shape from market to market. He said that:
“Lots of pilot projects are pushing forward, but in a very heterogeneous way. This is a very challenging scenario, because if we’re going to invest a lot of money we have to be sure that the solutions we create fit a wide variety of countries.”
Nevertheless, the company is moving forward with a number of autonomous technology efforts, including manufacturing parts for robotaxis, some of which are in use in the EZ10 autonomous shuttle from France-based company EasyMile.
In August 2019, Continental underscored its focus on electric vehicles when it announced that it would prioritize development of electric powertrains over internal combustion components. The move came a few months after the company appointed Dirk Abendroth, an auto industry veteran with expertise in electric powertrains and autonomous driving, as its technology head.
The German auto supplier is also teaming up with startups to expand its tech offering. In October 2020, Continental bought a minority stake in lidar startup AEye. The California-based company develops long-range lidar sensor technology that will complement Continental’s short-range lidar solutions. Frank Petznick, head of the advanced driver assistance systems business unit at Continental, said that this acquisition helps the company “facilitate automated driving at SAE Levels 3 or higher.”
In April 2017, Continental expanded its Silicon Valley research operations by opening an R&D lab in San Jose, California. The lab focuses on developing self-driving cars that can communicate with one another and with roadway infrastructure. Continental also said it would invest $300M to expand electric and hybrid vehicle technologies.
At the North American International Auto Show early in 2018, Continental executives discussed reorganizing the business and working to focus more on mobility and self-driving technology.
A few weeks later, Continental announced a new partnership with Nvidia to create self-driving vehicle systems, combining Continental’s automobile software engineering with Nvidia’s Drive platform and operating system. The goal is to develop an independent piece of autonomous technology that can then be marketed and sold to other automakers and grafted onto an existing vehicle to provide self-driving capabilities without the need for any complex, time-consuming, or expensive integration.
Read more: Reuters, SFGATE, Automotive News Europe
DAF, Daimler, Iveco, MAN, Scania, and Volvo complete autonomous truck ‘platooning’ trip
- Daimler to invest $570M in autonomous truck technology and create a dedicated organization for infrastructure R&D
- Main project involves self-driving truck “platoons” connected through wireless signals
- Started testing self-driving connected trucks on the road in Oregon in 2017
In April 2016, 6 convoys of truck “platoons” completed the first-ever cross-border trip of its kind. The experiment featured a dozen trucks from a diverse group of European brands — DAF, Daimler, Iveco, MAN, Scania, and Volvo — originating from various factories and converging in Rotterdam.
In the semi-autonomous “platooning” concept the group tested, multiple trucks controlled by a lead truck are connected through wireless signals, forming a train with one truck following another. This allows more trucks to be controlled by fewer people, maximizes efficiency, and decreases drag. However, the trucks featured in the test still required human drivers to be on board as a precaution.
Separately, Daimler has been testing its own autonomous trucks in Nevada since May 2015. The company announced it would be investing $570M in autonomous truck technology at CES 2018. Daimler has also created the Autonomous Technology Group (ATG), an organization focused specifically on developing “automated roadmaps” and other infrastructure initiatives to support the technology.
In March 2019, Daimler acquired the US-based self-driving company Torc Robotics and integrated it into ATG. The new partners have expanded the testing of autonomous trucks to New Mexico. Their goal is to bring SAE Level 4 vehicles to the roads by 2030. Daimler has also invested in Luminar, a lidar company, to gain access to vital sensors that will be a part of its autonomous trucks.
The Germany-based vehicle manufacturer also teamed up with Waymo in late 2020. Daimler will provide a customized truck chassis that will be integrated with Waymo’s self-driving system.
In February 2019, Daimler and fellow German automotive giant BMW announced that they would collaborate on autonomous driving technology, with the goal of getting the project ready by the mid-2020s. Mercedes-Benz head of automated driving Michael Hafner said of the collaboration:
“We have learned that the development of these systems is a bit like climbing a mountain. Taking the first few meters from the base station to the summit seems easy. But the closer you come to the goal, the thinner the air around you becomes, the more strength is required for each further step, and the more complex become the challenges you have to solve.”
In August 2019, it was reported that Audi was set to join this alliance between Daimler and BMW, although none of the 3 automakers confirmed it.
Note: the Volkswagen Group owns a controlling stake in both MAN and Scania. Daimler Trucks is a division of Daimler AG, which was also part of the group that bought Nokia’s mapping assets for $3.1B. Volvo Trucks is now a distinct company from the Geely-owned Volvo Car Group, mentioned above.
In September 2017, Daimler announced that it would be testing connected trucks in a platooning configuration on public roads in the US following approval from the Oregon Department of Transportation. Daimler showcased the trucks, which can make an emergency stop while loaded to 78,000 GVW (gross vehicle weight), at the Portland International Raceway in June 2018.
Read more: CNBC, Engadget, BBC, Forbes, Trucking-Rave
Didi Chuxing is testing robotaxi services, secured major investments
- Opened AI lab in Silicon Valley autonomous driving tech R&D in March 2017
- Demonstrated a working self-driving car in February 2018
- Received permission from California to undertake further public testing of its technology
- Plans to operate over 1M autonomous cars by 2030
After absorbing Uber’s China unit, ride-hailing giant Didi Chuxing is now following its US counterpart into self-driving research. In March 2017, Didi opened its own artificial intelligence lab in the heart of Silicon Valley, creating a distinct unit to drive its R&D for intelligent driving systems and AI-based security for transportation.
In August 2019, the company said it would start using self-driving vehicles to offer free rides in autonomous vehicles to its customers in a district in Shanghai city. Didi plans to eventually expand this service to the rest of Shanghai, Beijing, Shenzen, and its other global markets.
As of June 2020, customers can use the Didi app to hail self-driving cars for free in a designated test area in Shanghai. A safety officer is present in every vehicle. According to Meng Xing, Didi’s chief operating officer, “the robotaxis are to be deployed in places where ride-hailing drivers are less available.” Didi plans to have 1M autonomous vehicles on the road by 2030.
The company’s expansion plans followed a few days after it spun off its self-driving unit as an independent company in 2019. The new business’s goal is to “integrate the resources and technological advantages of Didi’s platform, continue to increase investment in R&D of core innovative technologies, and deepen collaboration with upstream and downstream auto industry partners.”
Didi Chuxing has already poached a number of noteworthy engineers for its new lab, including Charlie Miller (formerly of Uber’s autonomous vehicle security unit) and Jia Zhaoyin, a senior software engineer at Alphabet’s Waymo unit. Both engineers have since left Didi. The China-based company is also one of several companies to partner with Udacity to hire graduates from the startup’s self-driving “nanodegree” program.
In March 2017, the Udacity and Didi joint partnership announced a contest focused on self-driving technology and safety, where teams could win $100K for developing the best “Automated Safety and Awareness Processing Stack.”
Didi also has a 60% stake in a self-driving joint venture with Volkswagen, located in Shanghai.
In February 2018, Didi announced that it had demonstrated a working self-driving car for the first time, reporting that it had built the software for the vehicle and constructed the hardware in partnership with various carmakers and suppliers. In May 2018, Didi received permission from the California Department of Motor Vehicles to begin conducting further tests of its self-driving vehicles on public roads in the state. According to Didi Smart Transportation Initiatives vice president and chief scientist Henry Liu, the company has 40 autonomous vehicles.
In February 2020, Didi partnered with a China-based operating system technology company, ThunderSoft, to share technology and expertise on autonomous driving, advanced driver assistance systems, smart assistants, among other things.
Didi also secured over $500M in a round led by the SoftBank Vision Fund in May 2020. This is the single largest investment in China’s self-driving sector.
Read more: TechCrunch, Bloomberg, FT, Recode
FAW Group boasts the first certified autonomous truck in China
- Launched a robotaxi service with Baidu in 2019
- Developed a certified autonomous truck with Plus that will enter mass production in 2021
FAW Group, a China-based manufacturer of trucks, buses, and cars, is moving forward with several self-driving initiatives. Together with tech giant Baidu, FAW launched the Apollo Robotaxi service in China’s Hunan province in 2019. The trial involves 45 Level 4 cars built by FAW that will navigate over 80 miles of open roads. Baidu will provide autonomous driving software, and a safety driver is assigned to each vehicle.
In November 2020, FAW’s J7+ commercial truck, jointly developed with the startup Plus, passed a certification test conducted by the China Automotive Technology & Research Center. This means it can begin on-road operations. Mass production is set to start in 2021. FAW J7+ is the first intelligent truck that passed the certification test in China.
FAW Group has invested in autonomous tech startup Pony.ai as well. Pony.ai’s technology will be integrated into a vehicle platform used in the Red Flag electric vehicle brand. Similar technology will be deployed in FAW’s trucks, although further details about this strategic partnership weren’t disclosed.
Read more: Tech Wire Asia, Business Wire
Ford teams up with Argo AI to develop autonomous tech
- Plans to roll out autonomous vehicles by 2022
- Acquired AI startup Argo for $1B
- Partnering with Domino’s, Postmates, and Walmart on autonomous delivery pilots
In early 2015, Ford announced its “Smart Mobility Plan” to push the company forward in innovative areas including vehicle connectivity and autonomous cars. This plan culminated in the formation of Ford Smart Mobility LLC in March 2016, a new subsidiary focused on connectivity, autonomous vehicles, and mobility (e.g. car- and ride-sharing services).
As part of its 10-year autonomous vehicle plan, Ford is steadily increasing its fleet and currently has around 100 autonomous test vehicles. It has pioneered the testing of self-driving cars in environments including snowy weather and complete darkness.
The company has committed to roll out highly autonomous vehicles within pre-mapped, “geofenced” areas, including Austin, Texas, by 2021. In April 2019, then CEO Jim Hackett reaffirmed that timeline but stated that the vehicles’ applications would be “narrow,” and that the automotive industry “overestimated the arrival of autonomous vehicles.”
In a speech at Ford Field in Detroit in 2019, Hackett re-emphasized the enormity of what autonomous technology would mean, for the company and society as a whole:
“When we break through, it will change the way your toothpaste is delivered, logistics and ride structures and cities all get redesigned. I won’t be in charge of Ford when this is going on, but I see it clearly.”
The Covid-19 crisis, however, has forced Ford to delay the launch of its self-driving service to 2022. The company needed more time to assess financial losses sustained in 2020 and understand the effects of the pandemic on customer behavior.
In June 2019, Ford announced the opening of a new research center in Tel Aviv focused on self-driving technologies including sensors, in-vehicle monitoring, and cybersecurity. Israel is evolving into a major hub for self-driving technology, with Intel, Continental, Samsung, Daimler, and General Motors also making investments or setting up shop in the country.
Ford has aggressively pursued external investment and acquisition opportunities since H2’16, backing or acquiring a number of companies working in AI, lidar, and mapping. Its biggest move came in February 2017, when Ford announced that it would take a majority stake in AI startup Argo, investing $1B over the course of 5 years. Argo will operate with significant autonomy, becoming a de facto AI research (and recruiting) center for Ford. As of 2018, Ford is testing out Argo’s technology with its third-generation Fusion model sedan. Ford also has testing projects underway in Michigan, Miami, Pittsburgh, and Washington, DC. And in October 2020, Ford and Argo AI released its fourth-generation autonomous test car. The new vehicle is equipped with better sensors, sensor-cleaning technology, and an upgraded battery cooling system.
Additional Ford partners include Domino’s Pizza, with a partnership launched in August 2017 to develop a fleet of self-driving pizza delivery vehicles, and Postmates, to enable on-demand autonomous delivery. In November 2018, the company added Walmart to its list of autonomous delivery partners, with a grocery delivery pilot located in Miami-Dade County.
Read more: Bloomberg, The Verge, Wired, TechCruch
GM teams up with Honda and Lyft, invests billions in Cruise automation division
- Launched semi-autonomous Super Cruise in 2018 Cadillac CT6
- Received $2.25B from SoftBank to support autonomous work
- Plans to seek US regulatory approval to deploy Cruise Origin vehicles without pedals or steering wheel
General Motors made waves in 2016 with a series of aggressive moves within the tech sphere: it bought Sidecar’s assets, invested $500M in Lyft, and acquired autonomous tech startup Cruise Automation. In May 2019, GM announced an additional $1.15B investment in Cruise, bringing the operation’s post-money valuation up to $19B.
These deals have already borne fruit, with Cruise prototypes following Google in expanding testing into Arizona. In October 2018, GM announced an autonomous vehicle partnership with Honda that will see Honda invest $2.8B in GM’s Cruise unit over 12 years.
However, a full commercial self-driving operation is taking longer to materialize than GM initially hoped. The company planned to have an autonomous ride-hailing service in operation by 2019, but in July that year it announced the launch would be pushed back to allow for additional testing.
Cruise CEO Dan Ammann said of the delay:
“Our goal is to get there as soon as possible. We want that moment to come as quickly as we can. But everything that we do right now is going to be gated by safety. And that’s why we’re increasing our testing and validation mileage just to get to that point as rapidly as possible.”
Nonetheless, California’s Department of Motor Vehicles approved GM’s plan to pull human backup drivers from the vehicles it tests in San Francisco in 2020. This signals that the company is making steady progress in the self-driving field.
In January 2020, GM unveiled a self-driving 6-seater, Origin, calling it “the beginning of the future beyond the car.” The vehicle lacks a steering wheel, brake and accelerator pedals, a rear-view mirror, etc.
While the company did not reveal when the vehicle will be commercially available, it announced in October 2020 that it will soon seek regulatory approval to deploy several Origin vehicles for testing purposes. Origin will not be available for purchase by customers but will be used as a ride-hailing vehicle through GM’s Cruise service. The goal is to start producing the vehicle in GM’s Detroit factory by the beginning of 2022.
In its partnership with Lyft, GM planned to deploy thousands of self-driving Bolt electric vehicles beginning in 2018. However, this partnership was not exclusive, and Lyft and Waymo made a separate deal in May 2017 to collaborate on their own self-driving pilot projects. In June 2018, GM CEO Mary Barra told reporters that while GM still had a financial investment in Lyft, the two were no longer actively working together on any projects in the self-driving space.
Separately, GM has been developing its own semi-autonomous technology in-house, with its delayed Super Cruise finally launching in the 2018 Cadillac CT6. In February 2020, the company said that its autonomous driving unit now employed 1,800 employees, although 8% of this workforce was cut in May because of the Covid-19 crisis. Amid the crisis, GM also shut down its car-sharing service Maven and has delayed the relaunch of its Book by Cadillac luxury vehicle subscription service.
Read more: The Verge, Bloomberg, Cadillac, AP, Reuters
Honda to mass-produce Level 3 vehicles, offering semi-autonomous features in Civic
- Introduced semi-autonomous driver-assist on Civic models
- Partnering with GM on self-driving cars and batteries
- Launched its R&D Center X in April 2017
- Plans to mass-produce a Level 3 vehicle approved by the Japanese government
In 2018, Honda and GM announced a $2.8B, 12-year autonomous vehicle partnership. The companies also have a joint venture in place to develop hydrogen fuel cell systems by 2020. But GM eventually backed away from the fuel cell project, opting instead for battery-electric vehicles.
Honda has also received approval from California to test autonomous vehicles on public streets (with restrictions on the number of vehicles and the testing methods). Like Apple, the automaker is using the GoMentum Station proving ground, with 2,100 acres of testing area for its self-driving fleet.
In November 2020, Japan’s Ministry of Land, Infrastructure, Transport, and Tourism approved Honda’s plans to mass-produce its Level 3 Legend sedan. The car will be equipped with autonomous tech called Traffic Jam Pilot that can navigate through congested highways. The system will collect data on the vehicle’s movement. When drivers want to take back control over the vehicles, the software will first check whether they are looking at the road ahead. The launch is planned for March 2021.
Honda also introduced semi-autonomous ADAS (advanced-driver assistance systems) options on its entry-level Civic, offering lane-keeping, automatic braking, and adaptive cruise control functionality. These features have become ubiquitous on luxury models offered by brands like Tesla and Mercedes but are increasingly common at mass-market price points.
In April 2017, the company launched its R&D Center X, following Japanese competitor Toyota in establishing a dedicated AI research lab.
Honda was reported to be in talks with Alphabet’s Waymo to deploy Waymo’s self-driving system as early as 2016. By 2018, the deal had fallen through — allegedly due to Waymo’s reluctance to share its autonomous technology with the automaker.
Read more: Fortune, WSJ, Verge
Huawei partners with BYD, plans to launch low-cost lidar products for smart cars
- Partnered with Vodafone to work on cellular car-connecting technology
- Created own driverless car R&D team with 200+ engineers
- Test drove self-driving Porsche Panamera at MWC 2018
- Plans to launch low-cost lidar products
According to Huawei‘s chief strategist Dang Wenshuan, 70% of autonomous vehicles’ value will reside in information and communication technology rather than in the cars themselves. So it’s unsurprising that the telecommunications giant has shifted resources toward developing autonomous vehicles in a bid to be the company that creates — and captures — that value.
Huawei is working with automakers in Europe (Audi) and China (GAC Group, Beijing New Energy Automobile, and Changan Automobile) with the goal of launching autonomous cars as early as 2021.
American automakers are conspicuously absent from Huawei’s list of partners — likely a result of restrictions put in place by the US government to limit the sale of Huawei components in the US.
In November 2019, Huawei said that it was partnering with China-based mapping company NavInfo to use its mapping data for navigation in its self-driving cars. Like Tesla, the technology company also plans to integrate 5G technology into its driverless vehicles.
A year later, Huawei entered into another partnership with China-based electric car maker BYD. The company will reportedly use Huawei’s Kirin chips in its vehicles. BYD and Huawei have worked together in the past: in 2019, BYD’s monorails and energy vehicles were equipped with Huawei’s AI and 5G technologies.
The telecom giant is focusing on other autonomous technologies as well. It plans to launch low-cost lidar sensors at a maximum price of $500 per unit. Huawei has formed a dedicated team to run this project with the overall goal to make lidar affordable for all smart vehicles.
Huawei’s bid to expand beyond its home turf of telecommunications is several years in the making. In 2016, the company came out with a white paper detailing how mobile network operators could prove valuable in the connected car space. Some of these areas include smart parking, fleet management, data related to in-car entertainment, and LTE-based emergency services.
Reports also indicate that Huawei has assembled its own driverless car R&D team, with over 200 developers as of February 2017.
Huawei partnered with Vodafone to demonstrate some of its latest innovations at the 2017 Mobile World Congress in Barcelona, including cellular technology used to connect cars called Cellular V2X.
At Mobile World Congress 2018, Huawei unveiled a Porsche Panamera equipped with its Mate 10 Pro technology for object recognition. The technology uses a smartphone to guide the car. The car performs automatic object recognition and differentiates other cars from people and inanimate objects to determine its course of action. This demonstration, however, was seen as less of an ambitious push into the self-driving space and more of a way to showcase Huawei’s generalized AI capabilities.
Read more: Financial Times, Huawei, ComputerWeekly, RCRWirelessNews, The Verge
Hyundai focuses on affordable driver-assist technology, teams up with Yandex
- Focus on creating affordable driver-assist technology
- Investing $35B over a 5-year period in electric and autonomous driving
- Aiming to bring an autonomous vehicle to market by 2025
- Teaming up with the Russia-based tech giant Yandex to develop autonomous tech
- Invested in Israel-based technology firm Autotalks
After debuting a 2014 TV commercial that showed a convoy of cars outfitted with Hyundai’s driver-assistance tech, Hyundai was more conservative by September 2015. Its former European head of operations, Thomas Schmid, asserted that autonomous driving would come “by far not as quick as everyone says,” giving a timetable of 10 to 15 years. Nevertheless, the Korean motor group intensified its efforts in 2016, ramping up investments in AI and setting up a new business unit to develop “hyper-connected” and self-driving cars in the near future.
These efforts gained full steam in late 2019, when Hyundai announced plans to invest $35B over the next 5 years on its self-driving, electric, and connected mobility initiatives. The move came in line with South Korea’s plans to beat other countries to have commercially viable autonomous driving vehicles.
Following this announcement, Hyundai said that it would test 6 autonomous vehicles on the roads of Seoul, with plans to take this number to 15 by 2021. It also launched BotRide, a 3-month free self-driving taxi service in Irvine, California, in November 2019.
In addition, Hyundai is working with the Russia-based tech conglomerate Yandex on building a driverless car. In June 2020, Yandex announced it would buy 100 customized versions of Hyundai’s Sonata cars. Test drives will initially be carried out in Detroit. The duo is working on creating both a driverless car and a self-driving software system that could be sold to automakers and car-sharing startups.
Much like its automotive philosophy as a whole, Hyundai is striving for an affordable system that it can offer to mass-market buyers, showing off its Ioniq autonomous concept at CES 2017. The automaker also hired the former head of GM’s autonomous technology development to run its new Intelligent Safety Technology Center, a combined research body for Hyundai and its affiliate Kia.
In January 2018 at CES, Hyundai said it would begin testing an autonomous SUV, aiming to test the technology by 2021 and to go to market by 2025. The model would be powered by Hyundai’s partnership with Aurora, a startup working on autonomous technology founded by ex-executives from Uber, Tesla, and Google.
Later, Hyundai announced that its next-generation safety system being sent to production vehicles would give drivers the power to remove their hands from the steering wheel for up to 15 seconds at a time.
An investment in the Israel-based firm Autotalks followed in July 2018. Autotalks builds communications systems that help cars connect and convey information to one another, which Hyundai touted as a way to help build better detection and sensors in its cars.
Read more: Bloomberg
Jaguar Land Rover focuses on drive assist, partners with Waymo to build self-driving cars
- Plans to deploy a fleet of 100 self-driving research vehicles on the road in Britain
- Partnered with Waymo in 2018 to build a fleet of self-driving electric cars
- Supplies cars for Waymo’s fleet of ride-hailing vehicles
In March 2018, Jaguar Land Rover announced a new partnership with Waymo to build a fleet of electric, self-driving cars. By spring 2019, test vehicles began appearing on the public roads around Waymo headquarters in Mountain View, California.
Up to 20,000 new Jaguar I-PACE models will be built to serve in Waymo’s fleet for its ride-hailing transportation service, with production on the vehicles beginning in 2020. In May 2018, Jaguar announced it would begin working on off-road self-driving technology as well.
The Waymo partnership would have been unthinkable not that long ago. In June 2015, former Jaguar Land Rover (JLR) Director of Research and Technology Wolfgang Epple stated that autonomous vehicles would run counter to the brand’s philosophy, as the company doesn’t “consider customers cargo.” He asserted that the Tata-owned companies would instead favor advanced assistance features that would help drivers without taking full control from them.
However, in February 2016, JLR joined a $7.9M UK program to further autonomous driving R&D, aiming to gather data on driving habits and test vehicle communications technology. In July 2016, JLR formalized plans to deploy a fleet of at least 100 research vehicles over the next 4 years to test self-driving and connected car technology on roadways in Britain. The company has yet to provide an update on whether it achieved this goal.
Jaguar is also working on some technological innovations related to autonomous vehicles, including a system that projects a series of bars onto the road to notify nearby vehicles and pedestrians if the AV is turning, stopping, or starting. It’s one solution to a problem that has continued to perplex autonomous automakers: how to enable AVs to signal their movements to their surroundings. However, its usefulness may be limited in bright light or inclement weather.
The company has also developed software that reduces the severity of motion sickness in driverless cars. By optimizing braking, acceleration, and lane positioning, the system makes the ride more enjoyable. The software update, released in August 2020, will further refine the self-driving technology used by Jaguar and Land Rover models.
Read more: Engadget, TechCrunch, AutoNews
Magna joins Waymo’s stable of partners, also pursues lidar & autonomous tech
- After scrapping its partnership with Lyft, Magna starts working on assisted-driving tech, Level 1-3
- Teamed up with Waymo to build a self-driving vehicle manufacturing plant in Michigan
The Canada-based auto supplier giant has long been the subject of self-driving interest as the world’s largest automotive contract manufacturer. The company was dubbed the “Foxconn of cars” in an autonomous world, manufacturing vehicles designed by tech companies like Apple and Waymo.
However, Magna made a U-turn on fully autonomous tech in January 2020, when it scrapped its 2-year partnership with Lyft to develop self-driving technology. The Canadian company said that it would focus instead on developing assisted-driving technology, or the Level 1-3 market, where it sees more growth opportunity until 2025.
Magna and Lyft, in which the former has invested $200M, will continue to work together on other software and hardware related to self-driving tech.
In January 2019, Magna announced that it would team up with Waymo to build a factory to manufacture self-driving cars in southeast Michigan. The facility, up and running since summer 2019, is producing autonomous versions of the Chrysler Pacifica Hybrid minivan and Jaguar’s I-PACE electric SUV — all capable of Level 4 autonomous driving — with a production rate of several thousand vehicles per year. Most of these vehicles will be used by Waymo’s autonomous robotaxi fleet launched in Phoenix.
The company is already supplying legacy automakers with capabilities for advanced-driver assistance systems (ADAS). In 2016, Magna partnered with solid-state lidar company Innoviz to round out the sensor package for its self-driving system. The supplier has also tapped iPod visionary and Nest co-founder Tony Fadell for its tech advisory committee.
Read more: Clean Technica, Bloomberg
Mando tested a Level 4 autonomous vehicle, develops advanced lidar systems
- Successfully tested a Level 4 autonomous vehicle
- Producing radar solutions used by companies such as Mobileye
Mando Corporation, a major car parts supplier, successfully tested Level 4 autonomous tech in February 2019 with a vehicle called Hockey. The car navigated 2.7 km (1.67 miles) of public roads in the city of Pangyo, moving at a speed of 40 km/h (24 mph). Mando plans to commercialize this technology by 2021.
The South Korea-based company also produces fixed and rotating lidar systems, radar, and camera modules. The radar modules are used by Mobileye, an Israel-based provider of collision prevention solutions for self-driving cars.
Mando is especially well-known for its safety solutions that can be integrated into autonomous vehicles. For example, the company built a radar that tracks the driver’s heart rate, as well as a system that prevents parents from leaving their children inside the vehicle. And in October 2019, Mando unveiled several advanced braking, steering, and suspension systems.
Read more: PR Newswire
Microsoft pursues collaborative strategy with automakers on self-driving car development, offers Azure cloud to startups
- Supplying Azure cloud services to companies working on self-driving cars
- Working with Toyota on robotics, AI, and self-driving car development
- Supplying HoloLens technology to Volvo in self-driving partnership
- Supporting startups with free Azure credits and engineers
Microsoft is dipping its toes into AV research, though it’s later to the game than other tech giants.
Building off the capabilities and relationships of its Azure cloud computing platform, Microsoft is focusing its autonomous vehicle efforts primarily on “connected car platforms” — back-end networks tasked with crunching the data that cars produce. That data, in turn, informs a wide range of autonomous vehicle functions, including braking, advanced cruise control, and lane assist. Microsoft is also developing edge technologies that run complex AI models even when a vehicle is not online.
Smaller companies can benefit from Microsoft for Startups: Autonomous Driving program. They get free Azure credits and access to Microsoft engineers, which can provide crucial help when deploying and running complex AI workloads. London-based Wayve, which is building self-driving tech that is aimed at enabling autonomous vehicles to operate in 100 cities, is part of the program.
In March 2016, Microsoft and Toyota announced the expansion of their 5-year-old partnership to support Toyota’s research in robotics, AI, and AV development. As of June 2016, the company’s strategy focuses on providing automakers with technological assistance, as opposed to developing a car itself. Business from auto clients is now driving strong growth in Microsoft’s Azure cloud business.
Microsoft has also reportedly weighed taking a stake in the HERE high-definition mapping service, currently owned by BMW, Daimler, Volkswagen, Mitsubishi, and several other shareholders.
In 2017, Microsoft announced it would begin offering its Azure cloud services to companies using Baidu’s Apollo self-driving platform for autonomous projects. Then, in 2018, it became the data partner at the 330-acre American Center for Mobility facility in Michigan.
In December 2019, Microsoft partnered with Ericcson to allow the latter to build its connected vehicle cloud on top Microsoft’s connected vehicle platform. This would allow automakers to deploy self-driving tech in vehicles at scale.
Read more: ZDNet, WSJ, The Verge
Nissan/Renault tests self-driving taxis, looks to cut costs
- Alliance Intelligent Cloud partnership to deliver cloud-connected cars in ~200 markets
- Partnership with Waymo to bring self-driving cars and trucks to Japan and France
- Expanded self-driving technology testing in 2016
- Began partnership with DeNA to run self-driving taxi trials in Yokohama
In February 2020, Nissan achieved a major milestone in its self-driving journey: its battery-powered autonomous car Leaf successfully undertook the longest and most complex road journey in the UK, covering a total of 230 miles. The car was on self-drive mode 99% of the time, though 2 engineers were present in the car the entire journey.
In 2016 at the New York Auto Show, then chairman and CEO of Nissan and Renault Carlos Ghosn promised that the group would have 10 vehicles on sale by 2020 with “significant autonomous functionality.” Since 2017, Nissan has been testing an autonomous model on the roads of Tokyo. Nissan and Toyota also announced a joint effort to develop standardized “intelligent” maps, perhaps in response to German automakers’ acquisition of the mapping company HERE.
Nissan expanded the scope of its projects in 2016, with plans to test single-lane autonomous driving in Japan, as well as a wider range of experiments from its Future Lab mobility research center. In April 2017, Nissan joined Mobileye’s crowdsourced autonomous vehicle mapping effort. Mobileye also powers Nissan’s ProPilot system, which resembles Tesla’s Autopilot and controls the vehicle’s speed, lane positioning, and distance from other vehicles with minimal input from the driver. In 2019, ProPilot 2.0 received government approval in Japan.
In February 2018, Nissan announced that it would be working with the Japan-based online game and e-commerce company DeNA on a self-driving taxi trail, Easy Ride, which ran public trials in March 2018 in Yokohama, Japan. The service was booked via mobile app, and passengers were taken to their destinations in a modified Nissan Leaf electric car.
Renault-Nissan-Mitsubishi introduced the production release of “Alliance Intelligent Cloud” in March 2019. The system, built on Microsoft’s Connected Vehicle Platform, will deliver cloud-connected services in cars in most of the 200 markets served by the companies.
The Renault-Nissan-Mitsubishi “alliance” has partnerships with other major players as well. In 2018, they announced a multi-year deal with Google to put Android-powered infotainment systems in their vehicles, with the rollout to commence in 2021. In June 2019, the Google/Renault-Nissan relationship deepened with the announcement of a deal to bring driverless cars and trucks to France and Japan.
In May 2020, the conglomerate announced a new plan to cut costs and overcome a slump in car demand. Nissan will take charge of autonomous driving, while Renault will focus on electric vehicles and powertrains. Meanwhile, Mitsubishi will keep working on plug-in hybrids.
Read more: ZDNet, The Verge, Nissan, IB Times, Recode, Bloomberg
Download THE PDF
Nvidia develops its autonomous vehicle computer platform, partners with Mercedes-Benz
- Unveiled 8 teraflop computing platform designed for autonomous cars in 2016
- Unveiled plan to build self-driving trucks in 2017
- Partnering with Mercedes-Benz to deploy its Drive computing platform
At CES 2016, GPU and semiconductor company Nvidia unveiled the Nvidia Drive PX 2, the second generation of its platform expressly designed for autonomous cars. With 8 teraflops of processing power, the platform is designed for deep learning, sensor fusion, and computer vision applications — all key elements of a potential self-driving car.
Since then, Nvidia’s early bet led to automaker, supplier, and tech partnerships, from Baidu and Tesla to Bosch and Toyota. The company’s Xavier system-on-a-chip, built on its Volta architecture, promises improved processing power and efficiency.
Of note is Nividia’s partnership with truck maker Paccar to develop self-driving trucks, unveiled in March 2017. Nvidia CEO Jen-Hsun Huang has asserted that the effect of autonomy in logistics and transport will be a major opportunity for the company.
In June 2019, Volvo announced that it had selected Nvidia Drive as its autonomous driving platform for its planned fleet of autonomous trucks. The Volvo deal comes shortly after a similar deal with Continental.
Mercedes-Benz has taken a similar route. Instead of adding more electronic control units (ECUs) to run more technology, the company will use Nvidia’s Drive AGX Orin computer system-on-a-chip. The new software and hardware architecture will be deployed by the end of 2024 in all of Mercedes’ new vehicles.
Nvidia and Mercedes will also work on developing AI functionalities that correspond to Levels 2, 3, and 4 of vehicle autonomy. The new computing platform will theoretically allow Mercedes to introduce additional revenue streams. Drivers might, for instance, be able to purchase apps and subscription services using over-the-air updates.
Read more: ZDNet, Nvidia, TechCrunch
PSA Groupe finds startup partner to install self-driving tech for vehicles
- PSA’s self-driving car traveled 300+ kilometers without supervision between Paris and Amsterdam in 2016
- Partnered with nuTonomy in 2017 to install self-driving tech in Peugeot vehicles
In April 2016, France-based PSA Groupe (including Peugeot, Citroën, and DS) announced that two Citroën cars had driven without driver supervision from Paris to Amsterdam. The vehicles navigated over 300 km (186 miles) without supervision on authorized stretches of road, with PSA claiming the cars had achieved Level 3 automation in this mode. The “eyes off” mode is slated to arrive by 2021, while Level 2 semi-autonomous “hands off” modes are already available with models such as DS7 Crossback, DS 3 Crossback, and Peugeot 508 & 508 SW. These features, along with electric vehicles and new models, form the core of PSA’s broader Push to Pass growth strategy.
In May 2017, Peugeot announced a partnership with the MIT self-driving spinoff nuTonomy to install the startup’s self-driving systems in Peugeot 3008 vehicles.
PSA is also working with roads operator VINCI Autoroutes on testing autonomous cars as part of its “Autonomous Vehicle for All” (AVA) program. In 2019, the duo tested a vehicle’s ability to autonomously navigate a tollgate and operate at cruising speed. PSA also tested the performance of its self-driving technology in situations where roadworks alter the traffic, drivers don’t take back control, and the motorway ends.
Read more: PSA Groupe
Samsung enters the driverless car race, tests driver-assist technologies on roads
- Received approval from South Korea to publicly test self-driving cars in 2017
- Acquired Harman for $8B
Not to be outdone by its American competitors like Apple and Google, Korean electronics giant Samsung has thrown its hat into the autonomous vehicle ring.
In March 2017, the company completed its $8B purchase of Harman, a leading supplier of in-car technology and connected vehicle systems. Samsung plans to leverage the acquisition to shape the in-vehicle experience. Digital Cockpit 2020 was born out of this partnership. The infotainment system consists of 8 cameras and 8 displays. It uses 5G to link vehicles with home and office spaces, providing a personalized driving experience. From gesture control to touching buttons to face recognition, drivers can control the cockpit in many different ways.
In May 2017, Samsung secured a permit to begin testing self-driving cars on South Korea’s public roads. The tech giant’s self-driving cars are based on Hyundai vehicles equipped with cameras and sensors.
In January 2018 at CES, Samsung unveiled a new technology platform designed to help automobile manufacturers build customized autonomous vehicles. The first planned product in the platform is a camera that offers lane departure warnings, pedestrian and collision warnings, and adaptive cruise control. Cars testing the new platform, known as DRVLINE, are already being tested on the roads in both California and Korea. By the end of 2019, however, Samsung shifted its focus from developing the DRVLINE platform and pursuing full autonomy to advanced driver assistance tech.
Another major area of development for the company is advanced memory technology, which Samsung says is essential to “enable effective, high-speed driver feedback” and ensure the safety of autonomous vehicles. The company has developed a number of technologies in the space, including solutions for map data storage and data transfer.
In December 2018, Samsung announced that it would build mobile network infrastructure, including 5G and V2X networks, for K-City, a testing ground for autonomous vehicles and other smart city technologies located in Korea.
Read more: Samsung, Forbes, TechCrunch
SoftBank’s SB Drive venture pushes forward, teams up with Sensible 4 and focuses on driverless shuttles
- Working on a self-driving bus service in Japan
- Teamed up with Finland-based robotics company Sensible 4 to deploy driverless shuttles
SoftBank Group‘s SB Drive, an autonomous vehicle joint venture between the Japan-based giant and research company Advanced Smart Mobility, received a $4.4M investment from Yahoo Japan in March 2017. SB Drive focuses on advancing self-driving technology, especially related to public and community services such as buses.
SB Drive is conducting trials and testing the commercial viability of fixed-route buses for community public transportation, as well as autonomous truck-based freight delivery. In January 2020, the company said it planned to start operating a self-driving bus service on a fixed, 2.5-km long route in the Japanese town of Sakai. France-based autonomous and electric vehicle company Navya will provide the buses and SB Drive will provide the autonomous driving software for the buses.
SB Drive has smart mobility partnership agreements with 4 municipalities in Japan.
The Japan-based company has also inked a partnership with Finland-based robotics company Sensible 4 to focus on deployment of driverless shuttles in Europe and Japan. The companies will begin by testing SB Drive’s Dispatcher and Sensible 4’s GACHA shuttles on roads in Finland.
The Covid-19 pandemic forced Sensible 4 to scale down the testing. Instead of picking up passengers on a public transport route, GACHA shuttles were tested in a closed experiment. They weren’t taking passengers and were monitored remotely using SB’s system.
Read more: Deal Street Asia, TechCrunch, VentureBeat, Sensible 4
Tata Elxsi refines autonomous parking tech, focuses on autonomous vehicle security
- Driverless car platform Autonomai adopted by a leading OEM in 2017
- Launched improved smart parking solutions in June 2020
Tata Elxsi, a division of the Tata Group, showcased an autonomous parking valet technology in January 2015 in which the car understands where open spots are and uses sensors to park itself.
Since then, the company has continued to refine this technology. In June 2020, Tata Elxsi launched improved smart parking solutions. Park By Memory allows drivers to train their cars to park in a constrained space. Once the vehicle learns the maneuvers, it will follow their pattern in a similar situation in the future. The Parking Slot Detection is another tech built by Tata Elxsi. It helps drivers identify an empty parking spot and generates multiple paths to park the vehicle.
In August 2019, Tata Elxsi’s then CEO Madhukar Dev said that the company would further focus on assisted driving technology, as demand for fully autonomous tech was slowing. Dev explained the rationale behind the move:
“The need for full autonomy seems to have disappeared for the moment and what is coming in is assisted driving. So, some elements of our Autonomai are of interest and relevance to our customers, which is what we are now licensing.”
The company also put a priority on security, designing a central unit in the car with extensive security measures that govern internal and external automotive communication.
In January 2020, Tata Elxsi unveiled its concept autonomous car RoboTaxi at CES 2020. The taxi is built on the company’s middleware AV platform, Autonomai, and is fitted with AEye’s sensor system, iDAR.
In June 2017, the company licensed Autonomai to one of the top 5 OEMs. The software connects hardware (cameras, radars, etc.) with the AI and machine learning algorithms used to train AVs in complex driving scenarios.
Read more: Tata, Express Drives
Tesla got an early lead but struggles to meet self-driving expectations, clashes with industry coalition
- Released Autopilot, its semi-autonomous driver-assist technology, in 2014
- Autopilot Hardware 3 released in April 2019
- In 2016, announced that all Tesla vehicles on the road would be able to receive self-driving technology through a software update
- Three fatal crashes involving Autopilot have hurt Tesla’s self-driving reputation
- An industry coalition criticized Tesla for making unsubstantiated claims about its self-driving capabilities
In the public eye, EV manufacturer Tesla became an early leading banner-carrier for advanced driver assistance and self-driving technology. CEO Elon Musk has been particularly bullish on the field, announcing in 2015 that the technology behind fully autonomous vehicles was only “2 to 3 years away,” with another “1 to 5 years” needed for regulatory approvals.
Tesla pushed its Autopilot software update to properly equipped Model S vehicles in October 2015, enabling auto-steering, lane changing, and parking features. Tesla’s deployment strategy and messaging were criticized following a series of crashes and its first Autopilot-driven fatality in the summer of 2016, although the NHTSA’s official report was favorable towards Autopilot and did not find a safety defect.
Since the accident, Tesla and Mobileye severed ties, with the California automaker seeking to consolidate control over the development of its radar- and camera-based system (eschewing costly lidar sensors). According to former Uber CEO Travis Kalanick, Tesla has also rebuffed a partnership with Uber to collaborate on self-driving vehicles.
Starting in October 2016, all Tesla vehicles were built with Autopilot Hardware 2, a sensor and computing package the company said would enable “full self-driving” capabilities once its software matured. The system traded Mobileye’s EyeQ3 for Nvidia’s Drive PX2 platform and required an activation fee to unlock the full autonomy promised. Users reported poor performance during the initial rollout of Autopilot 2.0 software, although the system has improved with subsequent updates.
In March 2018, a man died when his Tesla ran into a highway barrier in California. After federal investigators learned the individual, Walter Huang, had previously complained about his Tesla’s Autopilot feature malfunctioning, they opened an investigation into Tesla’s self-driving technology. After Tesla made data from the car public in its defense, the US National Transportation Safety Board stripped Tesla of its role in the investigation into the crash, launching a feud and setting back expectations for Tesla’s self-driving car project.
In August 2018, Elon Musk announced the upcoming release of Autopilot Hardware 3. Model X and Model S vehicles equipped with Autopilot Hardware 3 went into production in March 2019. The system features custom chips manufactured by Tesla, replacing the Nvidia Drive platform.
Tesla offers two tiers of self-driving capability in its vehicles: Autopilot and Full Self-Driving. Full Self-Driving (FSD) costs an additional $10,000 and includes features like Summon and Navigate on Autopilot, which enables lane changes and interchanges. In October 2020, Tesla launched the Full Self-Driving beta, an update that initially went to a small group of customers. A full release is expected in the months ahead. Ford, Uber, Waymo, and other members of the Self-Driving Coalition for Safer Streets, an industry coalition, have criticized Musk’s messaging. They argue that Tesla’s vehicles aren’t truly autonomous because an active driver must always be ready to take over.
Despite the title, Tesla’s vehicles are only designated autonomy Level 2 by SAE, meaning they are capable of some autonomous maneuvers but are not considered fully autonomous. Development is currently underway on a next-generation chip that Musk claims will be “3 times better” than the current chip.
Musk also claimed in April 2019 that the US electric vehicle maker will be able to launch fully autonomous robotaxis by the end of 2020. But he acknowledged a year later that achieving this goal in a proposed timeframe isn’t going to happen.
Read more: TechCrunch, Forbes, Wired, Bloomberg, The Verge, The Washington Post
Toyota develops ‘Guardian Angel’ approach to autonomous driving, invests $2.8B
- Invested $22M in University of Michigan for robotics and AV research
- Focused on a “guardian angel” self-driving system, where a car will intervene when a human driver is about to make a mistake
- Invested $2.8B in Ford spinoff dedicated to autonomous technology
- Invested $400M into China-based self-driving startup Pony.ai in March 2020
Toyota has notably reversed from its 2014 claims that it would not develop a driverless car on public safety grounds. The Japan-based automaker announced at CES 2020 an ambitious plan to build a city to test its self-driving and smart technology as well as robot-assisted living.
According to Toyota’s CEO Akio Toyoda,
“With people, buildings, and vehicles all connected and communicating with each other through data and sensors, we will be able to test AI technology, in both the virtual and the physical world, maximizing its potential.”
The city is expected to house about 2,000 people, and will be located in the foothills of Mount Fuji. Work to start building the city is expected to start in 2021.
Toyota’s self-driving tech ambitions began in 2015, when it announced a $1B budget for autonomous driving research, establishing its advanced Toyota Research Institute (TRI) headed by Gill Pratt. Toyota has also hired professors and researchers from Stanford University, MIT, and the entire staff of the autonomous vehicle company Jaybridge Robotics. In April 2017, it also announced its third US university partnership with an automotive engineering stalwart, the University of Michigan.
In January 2018, Toyota revealed the latest iteration of its flagship autonomous vehicle. A few months later, it announced an investment of $2.8B in a new spin-off company, the Toyota Research Institute – Advanced Development (TRI-AD), designed to accelerate the company’s progress toward its self-imposed deadline for achieving testing of autonomous, electric cars.
TRI CEO Gill Pratt has been a vocal proponent of a “guardian angel” system, where the vehicle would monitor a driver’s inputs and intervene only when a human is about to make a dangerous mistake. TRI is also researching a second system (dubbed Chauffeur) targeted toward traditional Level 4 and 5 autonomy. The Chauffeur system would be primarily targeted at elderly people and people with disabilities.
TRI-AD has achieved several other successes as well. Together with Maxar, the institute managed to turn satellite photos into autonomous car maps. Although in a proof-of-concept stage, this breakthrough could make road-mapping faster and cheaper. Companies currently carry out mapping using a fleet of vehicles that are expensive and have to be redeployed each time roads are built or changed. Toyota is also working with Carmera, an HD maps and road intelligence startup, on researching whether dashcam recordings could be used to map roads.
As of January 2021, TRI-AD is to become a holding company called Woven Planet Holdings. It will own two subsidiaries, the first of which is Woven CORE, tasked with developing autonomous driving technologies. The second subsidiary is called Woven Alpha. It will explore innovative business opportunities beyond the automotive sector, including Toyota’s futuristic city.
Toyota is partnering closely with Nvidia on software to power its autonomous vehicles. Nvidia’s technology powers the full autonomous vehicle workflow for Toyota, from neural network training to testing, validation, and deployment.
“Our vision is to enable self-driving vehicles with the ultimate goal of reducing fatalities to zero, enabling smoother transportation, and providing mobility for all,” said TRI-AD CEO James Kuffner in a statement. “Our technology collaboration with Nvidia is important to realizing this vision. We believe large-scale simulation tools for software validation and testing are critical for automated driving systems.”
Kuffner said in December 2019 that Toyota will offer self-driving features first in commercial vehicles such as ride-sharing services, ambulatory services, before providing them in personal use vehicles.
The Japan-based company also infused $400M into China-based self-driving startup, Pony.ai, in March 2020 to deploy the latter’s autonomous driving tech in Toyota’s Lexus RX vehicles in Shanghai and Beijing.
In August 2019, Toyota inked a partnership with Suzuki to work on autonomous vehicle tech. As part of the deal, Toyota took a stake then worth ¥96B ($908M) in Suzuki, while Suzuki invested ¥48B ($454M) in Toyota.
Read more: TechCrunch, Wired, Wall Street Journal, CNET, CNN
Uber sells self-driving unit ATG to Aurora Innovation, ending its pursuit of autonomous vehicles
- Hired 40 engineers from Carnegie Mellon Robotics Lab to work on self-driving tech for Uber in Pittsburgh
- Has since faced legal troubles and criticism over its role in a fatal crash involving a self-driving Uber vehicle
- Conducted limited tests on San Francisco’s streets in early 2020
- Sold self-driving unit ATG to Aurora Innovation in December 2020
Self-driving cars initially had a big role to play in Uber‘s goal of becoming the “one-stop shop for the movement of people and powering local commerce around the world.” Its $7.3B self-driving unit, Advanced Technologies Group (ATG), accounted for a significant chunk of Uber’s valuation in April 2019.
ATG was spending $20M per month as of 2016, according to court documents that were unsealed in the lead-up to the company’s IPO in 2019. A major reason for why Uber was focused on self-driving is that it would help curb a major cost to the company: human drivers. Drivers account for nearly 80% of the company’s cost per mile for a ride, according to data by Frost & Sullivan.
The documents also reveal just how quickly Uber expected self-driving adoption to accelerate: it planned to have 75,000 autonomous vehicles in operation by 2019 and to have driverless taxi services operating in 13 cities by 2022.
The company fell short of that first milestone: in April 2019, it had just 250 autonomous cars on the road.
Uber has had a tumultuous relationship with autonomous vehicle development.
Former Uber CEO Travis Kalanick was a noted champion of his ride-sharing company embracing autonomous cars, with the technology representing both an existential threat and an opportunity for the startup to validate its stratospheric valuation. In 2016, he said, “What would happen if…we weren’t part of the autonomy thing? Then the future passes us by basically, in a very expeditious and efficient way.”
The company has made several moves in that direction, such as poaching nearly the entire Carnegie Mellon Robotics Lab (40 engineers) to work on the project in Pittsburgh. Uber also partnered with the University of Arizona to develop better mapping and optical safety technology.
In May 2016, Uber revealed its in-house autonomous prototypes for the first time, and the company acquired self-driving truck startup Otto later in the year. The acquisition embroiled Uber in legal controversy, with Waymo alleging that employees stole confidential files relating to its proprietary lidar sensors.
That lawsuit settled, as Uber agreed to give Alphabet $245M in equity. The company has also faced waves of negative press on sexual harassment and poor culture, in addition to deceiving authorities and ignoring autonomous testing regulations. The company has suffered from an exodus of both engineering and non-engineering talent. And the company has had numerous troubles with its self-driving cars and safety.
Just hours after Uber began self-driving trials in San Francisco, reports surfaced of a close call with a human-driven car. Subsequent reports alleged that Uber vehicles were self-driving into bike lanes throughout San Francisco without warning. Finally, in March 2018, a self-driving Uber vehicle was involved in a fatal crash in Arizona, striking and killing 49-year-old Elaine Herzberg. Uber was subsequently banned from testing autonomous vehicles on public roads in Arizona.
In the aftermath of the crash and the ensuing investigation, Uber suspended all of its self-driving car trials and pulled all its autonomous vehicles off the road. Around the same time, Uber announced the end of its self-driving truck program, Otto. However, in August 2018, Uber announced a $500M investment from Toyota and a plan to jointly work on autonomous vehicle development.
In June 2019, Uber unveiled a new self-driving car model, this time developed in partnership with Volvo. The model features more safety features and redundancy than the previous self-driving models Uber has developed — but it is also being designed to operate without a human driver behind the wheel. Uber completed limited testing in March 2020 on the streets of San Francisco.
Uber secured new capital for the venture as well. In April 2019, it was announced that the company had raised $1B in investments to fuel its autonomous driving efforts: $333M from the SoftBank Vision Fund and $667M from Toyota Motor Corp and Denso Corp. In a move toward reviving the testing of its autonomous vehicles, Uber also received licenses in February 2020 to resume its self-driving trials in California and to launch trials in Washington DC and Dallas.
However, amid growing losses and wider profitability concerns, Uber sold ATG to autonomous vehicle startup Aurora Innovation in December 2020. The deal, which includes Uber transferring its ATG equity to Aurora and a $400M investment in the combined company, marks an end to Uber’s pursuit of self-driving cars. Uber will reportedly maintain a 26% stake in the combined company, and as part of the deal, Aurora’s autonomous vehicles will one day be deployed on Uber’s ride-hailing platform.
Read more: TechCrunch, New York Times, Wall Street Journal, The Verge
Valeo partners to develop sensor technology
- Partnered with Mobileye in 2015 to work on an affordable self-driving car
- Unveiled technology at CES 2018 for identifying and analyzing individual passengers within a car
Like many other suppliers, Valeo has been working on its own advanced driver-assistance systems and self-driving systems, with the company’s innovation chief tapping autonomous driving as one of its main growth drivers.
The auto supplier showed off its eCruise4U automated concept at CES 2017, which was equipped with a connected camera, laser sensors, and a reconfigurable cockpit.
Valeo’s quarterly sales rose 22% in the first quarter of 2017, partly driven by the company’s push in ADAS technology alongside partner Mobileye. The two have an agreement dating back to 2015 to build an affordable robocar, combining Mobileye’s camera vision expertise with Valeo’s lidar units.
At CES 2018, Valeo demonstrated its system that identifies individual passengers within a car and can adjust the car’s interior according to the passengers’ preferences. At the same conference a year later, the company signed an agreement with Mobileye to develop a new safety standard for autonomous vehicles based on Mobileye’s mathematical safety model.
The same year, Valeo became the first company to test its autonomous car on the streets of its hometown, Paris.
Valeo’s team is based in Tuam, Ireland, and teamed up with Ireland-based software company Lero to create new, more durable autonomous vehicle sensor technology. Lero tests its sensors at the National University of Ireland at Galway, where it can rain as many as 225 days a year. The group hopes to develop smarter autonomous car technology that is more road-conditioned and weather-resistant.
In line with these plans, Valeo has developed sensor cleaning technology that washes cameras, lidars, and other optical sensors. Ensuring constant, high-quality video flow is vital for the safety of drivers in traditional and autonomous vehicles.
Read more: Valeo
Volkswagen plans to centralize vehicle software development
- Volkswagen began investing in autonomous driving initiative in 2016
- Unveiled self-driving shuttle concept car called the “Sedric” in 2017
- $7B joint venture with Ford announced in 2019
- Launched an internal organization called Car.Software to centralize software development
While the head of Volkswagen‘s self-driving unit recently admitted that full autonomy in vehicles in all conditions may never become a reality, the German company has been putting its full weight behind self-driving.
In October 2019, Volkswagen spun off its self-driving startup into a subsidiary, Volkswagen Autonomy, in a bid to “make autonomous driving market-ready.” This was followed by Volkswagen’s parent, Porsche SE, taking up a minority stake in self-driving tech startup Aeva. Volkswagen has also struck a deal to launch fully-functional autonomous van service in Doha, Qatar, by the end of 2022.
The automaker is further expanding its Volkswagen Autonomy initiative by hiring up to 100 new systems engineering experts for its existing Innovation and Engineering Center California. Additional offices are to be opened in China, as well, where the company has started testing autonomous vehicles in the city of Hefei.
VW is also eyeing autonomous trucks. Its heavy-truck unit, The Traton Group, acquired a minority stake in a self-driving truck company, TuSimple. Traton plans to equip its Scania trucks with autonomous tech and test it on roads in Sweden, Germany, and other European countries.
The road to Volkswagen’s self-driving ambitions started in 2015, when it revealed its V-Charge project. The plan was for a Volkswagen e-Golf equipped with sensors and 3D maps to find open parking spaces in a garage and park without human input. The company suggests that there will be a prototype for demonstration available within 4 years.
Former VW Group CEO Matthias Muller announced in March 2016 that the board had signed off on a huge autonomous driving initiative, claiming that the goal was to “[bring] these technologies to market faster than the competition.” The Group’s head of digitalization has said that self-driving cars will be “commonplace” by 2025.
The auto giant unveiled its new Moia brand in late 2016, joining other automakers in creating a unit dedicated to new mobility services like fleet-based shuttles and autonomous, on-demand transportation. Shortly thereafter, the company unveiled its “Sedric” concept, a fully autonomous shuttle-like vehicle designed for ride-hailing applications.
The Vizzion — the fourth in Volkswagen’s line of autonomous car concepts — was unveiled at the 2018 Geneva Motor Show. With no steering wheel or pedals, the vehicle represents Volkswagen’s vision of a fully self-driving vehicle that requires no human input to drive whatsoever.
In July 2019, Volkswagen announced that it had entered a joint venture with Ford to develop autonomous vehicle technology. The partnership will see the companies invest $7B in Argo AI, Ford’s autonomous driving platform, and will give Volkswagen an equal stake as Ford’s. Once complete, Argo AI will have the largest geographic reach of any autonomous driving technology to date.
Despite multiple partnerships, Volkswagen is keen to centralize the development of software technologies, including autonomous driving. To that end, the company launched an internal organization, called Car.Software, in June 2020. The unit is tasked with increasing VW’s share of its vehicles’ software to 60% by 2025, up from the less than 10% it currently owns.
Currently, software tools in Volkswagen vehicles are built by major IT companies and third-party providers, a common practice in the car industry. “Volkswagen wants to retain control of the entire vehicle architecture — that includes the electronics,” said Christian Senger, Car.Software’s CEO. “We cannot give third parties complete access to data in our vehicles.”
Read more: Forbes, The Verge, Reuters, CNET
Volvo Cars prioritizes safety in self-driving tech & perception with strategic partnerships
- Sought to build a reputation for safety in self-driving tech with IntelliSafe
- Lidar technology produced in collaboration with Luminar spots objects at 250 meters
- Plans to integrate Luminar’s lidar tech in select vehicles starting in 2022
The Geely-owned Volvo car brand has placed an emphasis on safety innovations when it comes to developing self-driving passenger vehicles. Volvo labeled its autonomous vehicle project “IntelliSafe,” setting a zero-fatality goal for when it fully rolls out autonomous features to the public.
That’s not the only goal Volvo has set for its autonomous efforts — in 2018, the company aimed for self-driving cars to make up one-third of its sales by 2025.
In June 2019, Volvo and ride-sharing company Uber announced that they had created “a production car capable of driving by itself.” The car in question is Volvo’s XC90 SUV, modified with Uber’s self-driving system.
Volvo’s statements about the autonomous XC90 have focused on the safety features, which include “several back-up systems for both steering and braking functions as well as battery back-up power,” according to a statement from the company.
Volvo has previously announced plans to give 100 Swedish customers early access to an autonomous XC90 by 2021 (with restrictions on when, where, and how the autonomous mode will be used).
(Note: Volvo Cars is now distinct from the independent Volvo Group brand, which focuses on heavy-duty commercial trucks, buses, and equipment.)
The company has stated that it will accept full liability when its vehicles are in autonomous mode and has announced plans to expand its pilot program to China and the United States. Volvo has followed rivals like BMW in setting 2021 as its target deployment date.
Volvo has other autonomous innovations in the works as well. In January 2017, the company announced a self-driving joint venture with Veoneer called Zenuity, which is focused on self-driving vehicle software. In June 2019, it was announced that Zenuity would provide self-driving software for Zhejiang Geely Holdings, the China-based auto company that acquired Volvo from Ford in 2010.
As of 2018, the joint venture is negotiating with 5 potential customers and aims to close at least one new deal through the rest of the year.
In June 2018, Volvo announced another partnership with lidar startup Luminar to work on both physical, car-mounted sensors and the software designed to process, label, and tag captured data. In November 2018, Luminar CEO and founder Austin Russell stated that the technology Luminar is developing with Volvo has the ability to detect objects at a distance of as much as 250 meters — significantly higher than the 30-40 meters most systems are capable of today.
If true, the technology would represent a substantial leap forward in tackling one of the most persistent challenges with self-driving vehicle technology: the ability to recognize — and react to — a constantly changing physical environment.
“Perception is the cornerstone to solving the autonomous drive problem in a safe way,” said Volvo R&D head Henrik Green in an interview with Automotive News Europe. “You need to see the object, understand what it is and, most importantly, predict the intent of that object.”
Taking the partnership with Luminar further, Volvo plans to integrate lidar and a perception stack into select vehicles, starting with the XC90, in 2022. Luminar’s lidar, along with autonomous driving software and other systems, will enable vehicles to run its highway pilot feature.
Read more: Auto News Europe, Forbes, Volvo, Wired, IB Times, Tech Crunch
Waymo (Alphabet) receives license for robotaxis in California, expands self-driving fleet and services
- Waymo One self-driving car service launched in December 2018
- Permission to transport passengers in Waymo robotaxis granted in California in 2019
- Surpassed 20M self-driven miles by Waymo autonomous vehicles in November 2017
- Purchased 62,000 new Chrysler minivans to increase size of Waymo self-driving fleet 100x
- Expanded partnership with Fiat Chrysler Automobiles in July 2020
The Google Self-Driving Car Project is one of the most iconic and tenured autonomous vehicle programs. Google hired ex-Hyundai and TrueCar exec John Krafcik to lead the program in September 2015. It later brought a legal lead on board, indicating an increased focus on commercialization under the new Alphabet structure.
The project formally became Waymo in December 2016. It expanded its testing beyond Mountain View and Austin to Kirkland, Washington, and Phoenix, Arizona.
In March 2020, it raised $2.3B in its first external funding round, with investors including Silver Lake, Andreessen Horowitz, and AutoNation. It also announced that it had accrued more than 20M miles on public roads and 10B simulated miles.
The company revealed a partnership with logistics company UPS in late January 2020 to test delivery of packages along a pre-set route in its self-driving vans in Arizona. Waymo has also said that it would bring its driverless vehicles to the streets of Texas and New Mexico.
The self-driving company has further expanded its footprint in the logistics sector by signing a deal with Fiat Chrysler Automobiles in July 2020. The duo will develop a range of autonomous vehicles designed for moving goods, including cargo vans. To start, Waymo’s self-driving tech will be integrated into FCA’s Ram ProMaster vans. The deal also covers other vehicles made by the Italian-American automaker, as well as those it might produce with France-based Groupe PSA if the planned merger of the two European giants goes through in 2021.
In April 2019, Krafcik painted a picture of the scope of Waymo’s ambitions for the future:
“Anything that has wheels and moves along the surface of the earth is something that we, in the future, could imagine being driven by Waymo.”
Waymo publicly revealed its custom self-driving hardware in February 2017, planning to sell an integrated hardware and software package. It opened signups for the first public tests of its customized Chrysler Pacifica minivans a couple of months later, quickly followed by its Lyft partnership.
In Q2’18, Waymo announced the purchase of 62,000 new Chrysler Pacifica minivans, increasing the size of its self-driving fleet by about 100x. In August 2018, the company launched a program to provide residents of the Phoenix area with rides to bus stops and train stations using the autonomous fleet.
Waymo launched Waymo One, its commercial self-driving ride-hailing service, in Phoenix, Arizona in December 2018. The program was initially piloted by a small group of approximately 400 Phoenix residents as part of Waymo’s early rider program.
The company received permission from the California Public Utilities Commission to operate its robotaxis on California roads in July 2019. The license comes with restrictions — Waymo can’t charge for rides, and vehicles must have a safety driver behind the wheel. Waymo is also not the first company to receive the public license: self-driving startup Zoox got its clearance to join the program 7 months earlier.
In November 2020, California authorized paid robotaxi services in the state. Waymo’s head of state policy Annabel Chang stated in response: “This long-awaited agency action will allow Waymo to bring our fully autonomous Waymo One ride-hailing service to our home state over time…and [we] look forward to putting our Waymo Driver to use in service to Californians.”
In October 2020, Waymo announced that members of Waymo One in the Phoenix area will be able to take fully driverless rides with no safety driver present. In the following months, non-members will also be able to use the service via Waymo’s smartphone app.
Waymo settled its high-profile lawsuit against Otto (and its parent company Uber) over the alleged misappropriation of Waymo’s proprietary lidar designs for $245M in 2018.
Read more: TechCrunch, Wired, TechCrunch
Yandex spins out its self-driving unit from JV with Uber into a separate company
- Separated its self-driving team from joint venture with Uber
- Clocked up over 4M autonomous testing miles
Yandex, the Russia-based tech conglomerate that runs a search engine and various other services, is keen on expanding into the autonomous driving sector.
The company’s self-driving unit was initially integrated into MLU BV, a ride-hailing venture with Uber. But in September 2020, Yandex announced that it was spinning out the unit into a separate company and investing $150M into it. Yandex will have a 73% stake in the spin-out, Uber 19%, and management and employees the remaining 8%. An autonomous delivery robot project, Yandex.Rover, is also being transferred to the new unit.
Yandex hopes that self-driving software will complement its robotaxi service, Yandex.Taxi, launched in 2018. “The goal is to develop a fully-fledged robo taxi service,” said Greg Abovsky, Yandex chief financial officer, “and that’s why we’re investing as much as we are.” The company also runs a food delivery business and other e-commerce ventures that could stand to benefit from autonomous driving solutions.
Yandex has assembled 130 vehicles that run tests across Russia, Israel, and the US. As of July 2020, the fleet has reportedly clocked up over 4M autonomous miles, positioning Yandex as one of the top three companies on this metric.
Read more: TechCrunch, The Moscow Times
Yutong has successfully tested driverless buses and smart road solutions
- Developed and successfully tested autonomous buses
- Invests heavily in smart road and terminal solutions
Chinese bus manufacturer Yutong has been researching driverless buses since 2012. The company claims to have successfully navigated a bus on an inner-city road in central China’s Henan Province. The bus can switch between manual and automatic mode.
Greyhound Australia started a 6-month trial of Yutong’s T12 coach in June 2018.
Yutong revealed a range of mobility solutions in September 2020. Its L4 autonomous driving buses were showcased operating in different scenarios. They were capable of autonomously navigating open roads, arriving at bus stops, and parking themselves. Yutong has also developed unmanned terminals and intelligent stops, and its smart roads technology ensures buses are given priority at intersections.
Read more: Yutong
ZF, Nvidia, Baidu partnering to build an autonomous car for China
- Invested in portfolio of self-driving tech companies, including Hella
- Partnered with Nvidia on integrating self-driving platform chipset into production vehicles
ZF Friedrichshafen AG, commonly abbreviated to ZF, is a major German auto parts manufacturer. It entered the autonomous vehicle space in 2015 with its $12.4B acquisition of TRW, a Michigan-based automated systems supplier.
Over the next few years, ZF began gradually acquiring and investing in its own portfolio of companies working on radar, camera, and lidar technology, including Hella, Astyz Communication & Sensors GmbH, Ibeo Automotive Systems GmbH, and doubleSlash Net-Business GmbH.
These acquisitions have paid off as ZF signed a number of deals for producing parts needed in self-driving vehicles. It’s partnering with self-driving truck startup TuSimple to co-develop sensor solutions — cameras, lidar, radars — used in autonomous vehicles. ZF has also signed a deal with US-based startup Aeva. The German supplier will produce lidar sensors that will be powered by Aeva’s algorithms and functionalities. These sensors are set to hit production by 2024.
At CES 2017, ZF’s then CEO Stefan Sommer announced that the company had teamed up with Nvidia to bring the chipmaker’s Drive PX2 AI Computing platform into production on automobiles. The system is expected to begin appearing in production vehicles in 2020, although there have been no recent updates on whether this has happened.
At CES 2018, Nvidia’s founder Jensen Huang announced that ZF systems technology would be used to produce a fully production-ready autonomous vehicle platform for China. The partnership combines ZF’s ProAI car computer with Baidu’s Apollo autonomous platform and Nvidia’s new Drive Xavier technology.
ZF made another bet in March 2019 on autonomous tech with its proposed takeover of rival Wabco. The $7B deal was approved by US and European regulators in January 2020, provided ZF sells Wabco’s steering components business in North America. The deal was successfully completed in May 2020.
Read more: Automotive News, Nvidia, TechCrunch, Reuters