Planning For A Digital Automotive Supply Chain

The amount of software in cars is growing rapidly, bringing with it both risks and opportunities.


At the heart of the very cool intelligent connected vehicles (ICVs) of tomorrow is sophisticated software with artificial intelligence and powerful silicon chips all working together. These technologies will transform the traditional automotive supply chain from a mechanical-driven world to a digital one, where the user experience, reliability, safety, and value are created from silicon and software. It will allow OEMs and Tier 1 suppliers to further differentiate, and do so quickly. It also introduces, if done poorly, additional risks to privacy, safety, and reputation.

The amount of electronics (chips and wiring) and software in new vehicles is steadily growing. Today the average vehicle is about half mechanical (body, chassis, tires, etc.) with about 15 percent software and 35 percent electronics. By 2025, the numbers are expected to be 35 percent mechanical, 25 percent software, and 40 percent electronics. This increase in software is significant and worth paying attention to.

There has been exponential growth in software lines of code in recent years. The number of software lines of code in the average vehicle today is about 150 million, compared with only 10 million software lines of code in 2010. This growth in software dramatically increases the complexity of different software stacks and electronics, and it increases the risks as well.

For the consumer, the immediate benefit of extra software will be found in the expanded and easy-to-use infotainment cockpit. New entertainment and productivity apps and services will deliver the extraordinary user experiences and open new opportunities across the entire automotive supply chain. Software will improve advanced driver-assistance systems (ADAS) so vehicles can be safer and more efficient on the road. And all this software can be updated over the air, reducing the overall amount of time each vehicle will spend in a service bay and reducing maintenance costs to owners over the lifetime of the vehicle.

These enhancements, however, cannot impact the overall functional safety of the vehicle. This will mean OEMs and their suppliers must meet additional requirements and comply with new industry standards. These requirements and standards must cover not just software, chips, and systems, but also the tools used to certify them for compliance.

The need for additional design and testing is real. Unlike most electronics and software applications, mistakes in automotive can be deadly. Vehicles have tremendous liability with their potential for the loss of life. The simple failure of a single part can result in high costs associated with recalls and damage to brand reputation. That’s why there needs to be greater emphasis on fault tolerance within automotive silicon and software.

New requirements needed for robust design must be newly formatted for intelligent connected vehicles. They must have a holistic view of system failure rate not only along the supply chain, but also throughout the automotive development lifecycle and the automotive production lifecycle. And these phases must be monitored continuously. This will require a paradigm shift.

Traditionally OEMs have acted more like system integrators and have not developed their own software or chips, relying instead on their supply chain to those. Creating their own robust electronics and software will require a fundamental change in the automotive development process as it stands today. OEMs will need to focus more on silicon and software. Additionally, traditional tech companies like Google, Apple, and Baidu will need to pay attention to the drive train and the more traditional aspects of automotive production (like chassis and body). This is where collaboration between OEMs and technology companies becomes valuable.

There are advantages to such a revamped system. Among them, faster time to production. By becoming more digital, OEMs and Tier 1s can take advantage of accelerated chip and software development. Virtual prototyping, for example, facilitates early integration and software testing and improves the overall quality. It also minimizes safety risks by thoroughly testing functional safety. This “shift-left” way of thinking about testing also reduces the overall cost of software and development and testing.

Leave a Reply

(Note: This name will be displayed publicly)