The next wave of disruption is here: Autonomous Vehicles and centralized computing will forever change the infrastructure of both vehicles and business models.
As we move into higher-level autonomous vehicles (L3-L4), there becomes a greater need for individual car components to communicate with one another.
True self-driving vehicles not only have to be able to brake, slow down, or detect external threats, but they also need to connect those functions to compute large, complex decisions—like how, if, and when it should slow down based on the detection of a potential threat on the road, and then triggering those functions accordingly.
To accomplish this, smaller individual semiconductors responsible for simple decisions, such as electronic steering, lights, windows, climate control, braking, etc. are being absorbed by larger, centralized semiconductors that create an interconnected system of various features, ultimately allowing the vehicle to develop accurate predictions and make critical decisions necessary to safely and properly drive the vehicle.
The consequence of moving into centralized compute systems is that Tier 1 suppliers who’ve traditionally supplied automakers with the technology behind individual vehicle functions could see a 25- to 40-percent loss in content contribution. Not to mention, the critical value-add and future source of innovation, semiconductor integration, and software development.
Evolution of the automotive vehicle architecture
Tier 1 suppliers currently supply around 75 percent of the electronic content inside a vehicle, and with these components comes the semiconductor that controls it.
In traditional vehicle architecture, each function of the vehicle—from energy management, display command, and climate control to steering, braking, and drivetrain—is powered by individual semiconductors that are developed and integrated by individual Tier 1 suppliers.
When we consider the dozens of parts provided by suppliers, it is the semiconductor and the ability to integrate that semiconductor with individual models and variants that provide the most value to automakers. It is also how most suppliers are able to differentiate from other supply companies and achieve a competitive advantage in the market.
Therefore, as suppliers begin to lose control of vehicle’s compute systems, they not only face a major content loss of up to 40 percent, but they also lose their key future value driver and could suffer eroding profit margin.
Who will own the centralized compute systems?
Currently, Tier 1 auto suppliers are playing a role in and contributing to the development of Level 1, 2 and 2+ Advanced Driver-Assistance Systems (ADAS) compute requirements. However, as we move into Level 3 and 4 Autonomous Vehicle (AV) technology and the major software and computer systems increasingly become the key differentiators between automotive brands. It is almost certain that carmakers, robo-taxi companies, and large semiconductor suppliers will fight to control the market (and the key source of innovation – semiconductor integration and software), creating fierce competition for Tier 1 suppliers, which could push them out of the competition entirely.
Traditional automotive suppliers simply don’t have the resources and ability to compete against top-tier suppliers, like APTIV, Denso, and Bosch, and large financially dominant semiconductor players, such as Nvidia, Intel, and Qualcomm who are able to scale up and serve the needs of these complex systems.
Even harder to compete against will be the automakers themselves, who will be likely to take ownership over the centralized compute and software systems responsible for the performance of their vehicles.
In traditional cars, it was the roar of the motor, the gas mileage, and the hum of the engine that set vehicles and brands apart. However, as battery-powered operating systems remove these elements, the new major differentiator will be the self-driving features and functions that are powered by the software and computers that control the vehicle’s autonomous driving capabilities.
Instead of horsepower or driver control, it will be the software and data that will create the feel, comfort, convenience, safety, performance, and cost of a vehicle, thus becoming central to the brand of the vehicle and the automaker.
Another reason OEMs won’t be likely to let go of the main compute system is because the value of the data created will simply be too high to pass off. The driver data collected from the software systems will provide valuable insight to automakers, insurance companies, and countless third parties, much similar to how Google and Apple have developed a business around individual data generated from mobile phones. A typical mobile phone generates 1 GB of data a day. Modern vehicles generate around 25 gigabytes of data every hour! Autonomous cars will generate even more – up to 3,600 gigabytes of data per hour, according to expert forecasts.
A centralized compute platform could also play a major role in updates, upgrades, and repairs. In Tesla’s software-defined vehicles, critical updates and upgrades can be sent to the vehicle via air, similar to updates to your phone or computer, which is one of the reasons Tesla has been hesitant to release control to a third-party supplier.
Because these computer systems will be so entwined with the performance, brand, and value of the automaker, it is almost certain that they will emerge as the primary leader in centralizing software technology.
Preparing for the evolution of AV computer systems
For years, industry experts have been ringing the alarm to the impact vehicle electrification would have on the industry, and now that suppliers are starting to see the realization of those predictions, they are also being hit with the arriving reality of autonomous vehicles and how this additional shift in vehicle architecture will further alienate traditional suppliers.
Tier 1 suppliers need to consider both the immediate impacts of electrification AND the quickly arriving effects of autonomous driving systems.
Automotive suppliers should be asking themselves now how centralized compute architecture will ultimately impact their business. Then, they need to ask themselves how they will be able to compete with top-tier semiconductor suppliers, or the OEM themselves, in the development of these compute and software platforms—or will they have to create a new vision and direction for their organization going forward?
Again, suppliers need to evaluate their long-term strategic plans as they are poised to face steep, virtually unbeatable competition when it comes to centralized computing systems.
Failing to recognize the realities of the changing vehicle architecture could result in a significant and sudden loss for suppliers. The only way to take advantage of or remain neutral in the emerging auto market is by addressing and planning for these scenarios now with a clearly defined long-term strategy plan.