The Road to Autonomous Driving at Fiat Chrysler Automobiles
As the world’s automakers commit billions to autonomous driving, FCA chooses to rely on the experts
In the race to leverage machine learning to develop autonomous vehicles, automakers like Fiat Chrysler Automobiles (FCA) are faced with many challenges. Chief among them are the scarcity of talent and the development cost. Engineers who have the machine learning skill set critical to autonomous vehicle development are hard to come by. Startups only a few months old but with an assembly of individuals with relevant talent have attracted billion-dollar investments [1]. Additionally, developing autonomous vehicles requires massive capital outlays to develop hardware and software and execute the millions of miles of road testing needed to collect data for system training and validation.
Faced with these challenges, many automakers have announced multi-billion-dollar investments in startups and homegrown research efforts. For example, General Motors purchased Cruise Automation for a billion dollars in 2016 [2]. Similarly, Ford invested a billion dollars in Argo AI in 2017 [1] and plans additional investments of $3 billion in autonomous efforts over the next five years [3]. Toyota announced in 2018 a $2.8 billion investment to create new subsidiary focused on developing software for self-driving vehicles [4].
FCA, however, has taken a different tact. Rather than investing billions in its own proprietary efforts, FCA has chosen a more diversified and lower risk approach by working with outside partners. Additionally, it has split its efforts into short- and long-term objectives.
In the near term, FCA is working to put a Level 2 system into production with supplier, Aptiv, by 2020 [5]. Level 2 or “Partial Driving Automation” systems can drive the vehicle, but only in certain situations and only under the continuous supervision of the driver [6]. For Level 3 systems, FCA has entered a consortium with BMW, Intel, and Mobileye [7] with the goal of going to market in 2021 [8]. Level 3 or “Conditional Driving Automation” systems do not require human supervision but do require that a human be able to intervene if the vehicle encounters a situation it cannot handle [6].
In the longer-term, FCA has partnered with Waymo on Level 4 systems [5]. Such systems are termed “High Driving Automation” and are capable of driving under limited conditions such as in a geofenced area without need for a backup driver [6]. Waymo is the world leader in this technology and is running an autonomous car service pilot in Chandler, AZ with retrofitted Chrysler Pacifica Hybrids. Additionally, FCA and Waymo are in talks to develop a self-driving car for personal use [9].
This three-pronged approach grants FCA the ability to learn from many partners, bear only a portion of development costs, and reduce risk by simultaneously betting on several approaches to autonomous driving. Additionally, it allows FCA to continue to focus on its core competencies while leveraging industry-leading technology from its partners. However, there are risks that FCA’s management team must address.
First, in relying on partners to develop automated driving systems, FCA risks falling behind in machine learning competency more generally, which could negatively impact its ability to deliver other innovative features and services to customers (e.g. passive vehicle unlock using facial recognition). Management should therefore focus on creating this expertise in-house so that FCA can keep up with trends and deliver differentiated user experiences.
FCA’s management team must also focus on recreating a customer value proposition in the age of the autonomous vehicle. Traditionally, features such as vehicle performance and driving dynamics have been key differentiators and contributors to brand identity. But If a car is rarely or never driven manually, these attributes will most likely lose relevance for customers. Management should therefore break the pattern of delivering marginal improvements to mature products and instead focus on a blank-sheet rethinking of what a vehicle will be and how it will provide value to its user once it can drive itself. This will require a transition from linear development processes to a broad exploration of this new space using divergent and convergent thinking followed by development and testing of concepts. FCA has a key advantage in being positioned close to Waymo and it should take full advantage of the opportunity to carefully study the ways in which customers interact with and experience the autonomous vehicles that are on the road today.
Moving forward, some questions remain. Will it be possible for automakers to avoid the commoditization of the automobile as autonomous vehicles become prevalent? Will vehicle sales for personal use be more than a niche market in ten years? Will autonomous systems be differentiated in ten years or will the technology itself be commoditized?
(751 words)
References
[1] T. Higgins, “Ford Acquires Majority Ownership of Self-Driving Car Startup Argo AI,” The Wall Street Journal, 10 February 2017. [Online]. Available: https://www.wsj.com/articles/ford-acquires-majority-ownership-of-self-driving-car-startup-argo-ai-1486756594. [Accessed November 2018].
[2] M. Ramsey and G. Nagesh, “GM Gives Its Self-Driving Effort a Push,” The Wall Street Journal, 11 March 2016. [Online]. Available: https://www.wsj.com/articles/gm-to-acquire-autonomous-vehicle-technology-developer-1457704950?mod=article_inline. [Accessed November 2018].
[3] K. Naughton, “Ford Is Far From First in Driverless Vehicles—and Investors Want In,” Bloomberg, 19 August 2018. [Online]. Available: https://www.bloomberg.com/news/articles/2018-08-09/ford-is-far-from-first-in-driverless-vehicles-and-investors-want-in. [Accessed November 2018].
[4] J. Lippert, “Toyota Adding $2.8 Billion to Self-Driving Software Outlays,” Bloomberg, 4 March 2018. [Online]. Available: https://www.bloomberg.com/news/articles/2018-03-05/toyota-adds-2-8-billion-to-software-push-for-self-driving-cars. [Accessed November 2018].
[5] “FCA Capital Markets Day: A Technology Update on Autonomous Driving and Connectivity,” Fiat Chrysler Automobiles N.V., 1 June 2018. [Online]. Available: https://www.fcagroup.com/capitalmarketsday/Presentations/FCA%20June%201%202018%20CMD_Technology%20Update.pdf. [Accessed November 2018].
[6] SAE International Surface Vehicle Recommended Practice, Taxonomy and Definitions for Terms Related to Driving Automation Systems, SAE Standard J3016, 2016.
[7] B. Gritzinger, “FCA Joins Global Autonomous Vehicle Development Team,” WardsAuto, 16 August 2017. [Online]. Available: https://www.wardsauto.com/technology/fca-joins-global-autonomous-vehicle-development-team. [Accessed 2018 November].
[8] “FCA Capital Markets Day: Jeep Brand,” Fiat Chrysler Automobiles N.V., 1 June 2018. [Online]. Available: https://www.fcagroup.com/capitalmarketsday/Presentations/FCA%20June%201%202018%20CMD_Jeep%20Brand.pdf. [Accessed November 2018].
[9] T. Randall and M. Bergen, “Waymo’s Self-Driving Cars Are Near: Meet the Teen Who Rides One Every Day,” Bloomberg, 31 July 2018. [Online]. Available: https://www.bloomberg.com/news/features/2018-07-31/inside-the-life-of-waymo-s-driverless-test-family. [Accessed November 2018].
Smart of FCA to double down on its strengths and recognize that it needed to partner with industry experts to remain competitive in the race to self-driving. I agree that FCA risks further increasing their ML knowledge gap by choosing not to invest in in-house development. I think it’s recognition that they’re too far behind to compete with the leaders in the space. The industry is really “adapt or die” – now more so than ever. To that end, this was the only option left for FCA.
I think the commoditization of the automobile is already underway and it will be difficult for OEMs to avoid. FCA’s positioning in the market and brand strategy may help them stave off this trend better than some of it’s competitors, but the risk is a real one.
While FCA’s strategy of relying on partners to help develop AV technologies in this time when its competitors are spending billions of dollars (on technologies that may be fruitless) may be considered a smart one, I believe that they absolutely risk being left behind. In an automotive industry which is changing more quickly than it has since its inception a century ago, by not building internal know-how in these technologies and best practices, you risk becoming obsolete.