Alphabet’s Waymo Partners with the Renault-Nissan Alliance on Autonomous Vehicle Deployment

In June 2019, the Renault-Nissan alliance issued a short press release announcing that it would be partnering exclusively with Alphabet’s Waymoto “explore all aspects of driverless mobility services for passengers and deliveries in France and Japan.” The latest Waymo announcement comes as the industry at large undergoes some major shifts in partnerships for autonomous vehicle software and hardware development. In the face of cost pressures and delayed implementation deadlines, Original Equipment Manufacturers (OEMs) are reevaluating and rationalizing their level of spending on autonomous driving development, either by pooling their in-house efforts with competitors or shifting more development to dedicated Autonomous Vehicle (AV) suppliers.

This spending rationalization by OEMs is having an impact on many autonomous vehicle startups, and the number of vendors in the ecosystem is unsustainable. Shortly after the Waymo announcement Apple acquired Drive.ai, which was once valued at US$250 million and on the verge of closing down shortly before its rescue. 

This section provides a summary of the new partnerships that have emerged over the last 12 months:

• BMW and Mercedes have announced that they are pooling their resources on autonomous technology development, following an earlier announcement of a partnership focused on the joint deployment of shared mobility services.
• Honda invested US$2.75 billion in GM’s Cruise, comprising a US$750 million equity investment, with the remainder to be invested the development of robotaxis over the course of the next 12 years.
• Volkswagen (VW) ended its relationship with Aurora, following an allegedly refused acquisition offer, and has since engaged in negotiations with Ford and Argo AI regarding the possibility of joint-investment in the development of a fully autonomous platform.
• In the meantime, Aurora has secured further investment from Hyundai, and has entered a partnership with Fiat Chrysler Automobiles (FCA) to develop autonomous commercial vehicles.
• Toyota has deepened its relationship with NVIDIA, opting to use NVIDIA’s hardware, software engines, tooling, and simulation solutions while maintaining its close ties to Uber.

There is a clear pattern emerging around the entry of robotaxi startups into the automotive supply chain, with those OEMs that have lagged behind in their autonomous software development rapidly committing themselves to dedicated partners. Waymo’s first OEM partnerships were established with FCA and Jaguar Land Rover (JLR). Both OEMs have been lackluster in their own autonomous vehicle technology development, and offered Waymo the ability to offer both mass market and premium mobility services. In both of these engagements, FCA and JLR behaved more as a supplier of the “box on wheels” to Waymo than as a development partner. Questions remain as to how, if at all, this latest Waymo partnership will differ from the previous FCA and JLR engagements.

The press release makes no mention of any technology development aspect to the relationship, with the exclusive partnership focusing more on the exploration of legal and market opportunities in deploying driverless ridehailing and delivery services in France and Japan. While it may be far too late for Renault, Nissan, FCA, JLR, Hyundai, and Honda to develop their own, independent autonomous technologies, there is clearly a desire to engage with partners that are willing to share development, rather than simply relegate OEMs to an effective chassis supplier role. This is likely one of the major factors that has driven VW from Aurora and towards Argo AI, whose more flexible approach will likely give VW greater ownership of the enabling autonomous technology.  

In the case of Renault-Nissan, the announced partnership with a major driverless technology developer represents an important “win” for the troubled alliance, which has been under considerable strain since the sudden departure of former Nissan CEO Carlos Ghosn late last year and the derailed merger attempt between Renault and FCA.

With robotaxi startups keen to establish a foothold with OEM partners in order to survive the forthcoming consolidation, OEMs must consider how best to select and engage with partners in a way that best fulfills their autonomous vehicle objectives. The advantages of partnering with a third-party specialist go beyond cost distribution, and include the ability for more concentrated intermediaries to train their neural networks based on the datasets coming from multiple OEMs, and greater expertise and experience with artificial intelligence. Below are some recommendations for how OEMs should choose and manage their relationships with dedicated robotaxi suppliers:

• Don’t Throw Good Money after Bad:Most OEMs have some level of existing investment in autonomous vehicle technology. There is always a temptation to maintain in-house investment in what many OEMs view as a differentiating technology, particularly in view of the vast sums already invested. However, the time has come for the OEM laggards in AV development to engage with a partner that can rapidly accelerate their position in the AV development race, or risk being left behind competitors. Certain AV technology players such as NVIDIA and AImotive have developed a full software stack, but are also flexible on how their customers incorporate their own software engines/modules for perception, motion planning, etc. alongside their own software products. For NVIIDA, this includes providing software libraries, tools, and collocated engineering teams to enable customers to leverage NVIDIA’s existing software stack while still developing differentiating software unique to specific customers.
• Pick One Partner for Level 4/5 and Scale Back:Many OEMs have more than autonomous technology development partner, and some have different partners to address different goals. For example, BMW and Daimler are now cooperating on highway automation but have engaged with completely different partners for their Level 4/Level 5 automation projects, with BMW developing an alliance with Intel Mobileye and Mercedes opting for a Bosch/NVIDIA-based platform. Therefore, both OEMs risk deploying semi- and fully autonomous platforms based on different platforms, requiring them to manufacture, deploy and maintain two separate hardware and software stacks. Toyota is also in a similar position, working with both NVIDIA and Uber on autonomous software development with different target applications.

A far more efficient policy would be to choose an autonomous technology partner that can provide a hardware or software stack (or a combination of both) targeted at enabling the ultimate Level 5 robotaxi application, scaling back to target Level 2+ supervised driving in the meantime to provide short term revenue. How should OEMs therefore engage with the likes of Waymo, Aurora, FiveAI, etc., which are committed to enabling only the fully driverless robotaxi use case? OEMs should be wary of subsidizing the continued development of such companies without first investigating how the software assets of the robotaxi company can be repurposed in the short to medium term.