In the Internet of Things — or the “Intelligence of Things,” which is a more apt description — the automotive segment has become a leading center of innovation. Only healthcare offers the same ability to leverage advances in sensors, processing, connectivity and artificial intelligence technology to advance an industry while simultaneously improving society.
As a result, the entire electronics and automotive ecosystem is racing to develop self-driving/autonomous vehicles. However, changing an established industry that traditionally has had long product development and life cycles does not come without challenges.
Recent panels at the NXP Connects conference in San Jose highlighted some of these challenges.The NXP Connects conference brought together many technology companies, automotive equipment manufacturers, and some new and traditional automotive vendors to discuss the future of the automotive industry. The first thing that stood out was the stark separation between new vendors and old vendors — or Silicon Valley (new) versus Detroit (old).
While the geographic associations are not completely accurate, especially with both camps spread across the globe, the discussions did highlight some of the differences between the two camps.
New Technologies, New Attitudes
The first difference between the new and the old is the mindset that drives how these companies do business. The traditional automotive ecosystem comprises technology vendors, systems vendors, automotive manufacturers and dealers in what is a very defined value chain with very defined timelines. A new vehicle design cycle typically takes about five years.
The traditional automotive value chain for more than a century has been defined by the concepts of dividing tasks along lines of expertise, and ensuring that everyone in the value chain can achieve a reasonable profit.
On the other side you have the new entrants, ranging from semiconductor startups to new car companies that have a very aggressive attitude of “charge ahead and worry about the consequences later.”
Companies like Uber and Tesla exemplify this mentality as seen by their decisions to introduce new technology before it is fully qualified, and often before the infrastructure, regulations and market are ready for it. Uber even used its technology to avoid regulations.
These startups are also willing to break the value chain. Tesla, for example, is developing everything from the AI processor through the rest of the car. In addition, Tesla sells vehicles directly to consumers through storefronts rather than through traditional dealers.
This battle already is breaking down the traditional value chain of the automotive segment and adding uncertainty to the market. On top of this battle, the industry is rapidly changing around the technology.
With the goal of achieving full driving autonomy, the automotive industry is pushing technology advances. Sensors are rapidly becoming smaller while increasing in fidelity; processor architectures are morphing to meet the power and performance requirements of mobile AI processing; and the wireless industry is moving to 5G to provide low latency cloud connectivity.
These moves are forcing the automotive industry to develop more flexible designs at a faster cadence, and to adapt to the technology as quickly as it changes.
At the same time, the automotive industry has been moving up the SAE levels of autonomy — level zero being no autonomous functions and level 5 being full autonomy — at a pace that may be beyond the infrastructure, laws and regulations to support it.
If that weren’t enough, the industry also has been adapting to the electrification of vehicles and changes in business models around concepts such as ride-sharing and vehicle-sharing. (For more detail on this economic and societal shift refer to Will the Sharing Economy Kill Personal Ownership?)
Long Period of Adaptation
The result of all this change is uncertainty as to how the automotive industry will look in the future, and what business models to adopt. The traditional value chain already is breaking down.
Chip vendors like NXP, Renesas and ST Microelectronics once supplied automotive system suppliers — typically referred to as the “tier ones” — like Bosch, Continental, Denso, Delphi and Magna. With the entrance of companies like Nvidia and Intel, the introduction of AI technology, and the complexity of the systems, chip providers now are supplying complete boards and systems, like the Nvidia Drive PX2 or the NXP Blue Box.
At the same time, many original equipment manufacturers have been developing their own infotainment and control platforms to be more vertically integrated, like Apple, and to ensure that they are not beholden to the chip/system vendors, or to major software vendors like Amazon, Google and Apple, which now dominate user interfaces and Internet services.
For fear of being locked out of the market, some tier ones have been developing full vehicle platforms. In other words, it’s unclear what the value chain will look like in the future.
In addition to the value chain confusion, it is not clear how revenue will be generated for the end consumer. With the electrification and automation of vehicles, the technology content has been increasing exponentially, which is increasing the cost to manufacture and maintain these vehicles. The increase in costs, combined with the changes in society, may push automotive OEMs to become car service providers.
Many of these issues were brought to light in the discussions at the NXP Connects conference, and some of the companies expressed growing concern about how and when they will make money. Many of the technology and automotive startups are banking on the transition to autonomous electric vehicles, but dates for the rollout of such vehicles continue to be pushed out.
Even the leading electric vehicle vendor, Tesla, has been struggling to reach critical mass and may face financial challenges before the market and consumers are prepared to make a full switch to electric or autonomous vehicles. Changes in regulations in China and Europe may push for faster adoption, but there may be too many companies fighting for a small portion of the market.
Unfortunately, those pushing for change in the market are the least likely to predict with any accuracy when the market will change, because their outlook is almost always too optimistic. Change is coming — but over the next few decades, not years.
It typically takes two to three generations of a technology before it is ready for mass adoption because the technology, ecosystems, and business or usage models around the technology need to adapt. We have seen this in everything from DVD players to smartphones.
If this holds true for the automotive industry, then electric vehicles will be hitting their growth curve in the next few years, but it likely will be closer to the end of the next decade before mass adoption of autonomous vehicles begins. There will be opportunities, but only for a select few who can bear the cost.
More importantly, it’s unlikely that the traditional automotive value chain will be replaced completely. Due to the expertise in each segment of the value chain, it will still exist in some fashion.
The most likely outcome is that you see closer partnerships between technology providers, system vendors and automotive OEMs. In terms of the business model, the most likely outcome is that automotive OEMs become service providers as a larger portion of consumers opt to lease, rent or share a vehicle instead of choosing traditional private ownership.
This also means that a good portion of the downstream services, such as dealerships, rental agencies, insurance and mechanics, will become obsolete.
The automotive industry is changing with technology and society, but it is not changing overnight. Companies that are part of the traditional value chain need to adapt. Startups hoping to break into the market need to be prepared for a long transition period while the market develops.