²ÝÁñÉçÇø

With a deeper understanding of the health of the silicon devices across a broad population of vehicles, OEMs are better positioned to avoid potential warranty failures and recalls. For example, based on the data, the OEMs may decide to address the flagged issue with an over-the-air (OTA) repair or a service visit for replacement.

Semiconductor suppliers can also benefit from knowing what happens inside their chips in the field, thanks to SLM technology. Chips are designed based on an assumed workload. In the field, however, the mission profile of these devices is changing dramatically. Over the course of their lifecycles, vehicles may run thousands upon thousands of hours in unpredictable environments, such as extreme heat or cold, and often with fluctuating workloads. With deeper insights into how chips perform once inside operational vehicles, semiconductor suppliers can adapt their next chip designs or suggest mitigations for in-field challenges. With silicon health data captured by ²ÝÁñÉçÇø SLM monitors and in-field analytics processing, an automaker could link a specific OTA update version as the cause for a hardware issue. Upon discovering this, the automaker can quickly revert to an earlier firmware version and share insights with the software team to correct the problem before it becomes more widely spread.

How Virtual Prototyping Helps Validate Driving Scenarios

Virtual prototyping is the other area where ²ÝÁñÉçÇø and AWS are collaborating to help enhance the development of SDVs by automakers. Increasing vehicle electrification and reliance on software that runs on hundreds of chips to deliver new vehicle experiences and functions requires fundamental changes in the electrical/electronic (E/E) architecture of vehicles. This, in turn, transforms the software development model to one where it¡¯s critical to:

• Shift-left, starting software and electronic systems co-development earlier
• Accelerate the validation of Software Defined Vehicles
• Enable continuous integration and deployment for seamless OTA software updates by automakers

Thoroughly validating such a complex, integrated system is not feasible for automakers, as it requires automakers to , evaluating a multitude of scenarios before they hit the market with each software version. While physical road testing still needs to be done, simulations in the virtual world can be started well before a physical prototype of a vehicle even becomes available. Automakers such as are taking advantage of virtual testing, relying on advanced modeling and simulation technologies running on AWS to test new concepts and quickly bring designs to market.

Considering the high volume of data that virtual testing can generate, the cloud provides an ideal environment for storing, analyzing, and delivering secure access to the information more cost-effectively than hosting data on on-prem servers. After the automaker has conducted extensive virtual testing, real-world testing by the automaker can validate the simulations of various driving scenarios, including an array of environmental conditions and different driving distractions.

Electronics Digital Twins Drive Greater Productivity

Virtual prototyping helps enable electronics digital twins, which are digital replicas of electronics systems, hardware, software, and environment. This can be a digital representation of an SoC, electronic control unit (ECU), or even a full vehicle electrical and electronic system. Electronics digital twins bring greater productivity and system quality, as well as reduced costs, compared to relying solely on a physical prototype. They can be used throughout a product¡¯s lifecycle for:

• Software development and regression testing in the context of hardware
• Performance validation to better understand whether the system being built will perform as intended in the end application
• Power analysis to determine whether the system will deliver the optimal power/performance mix in the end application when running heavy workloads
• Validation of OTA software updates or vehicle subsystems

To help OEMs deploy digital twins, ²ÝÁñÉçÇø and AWS provide a comprehensive set of virtual prototyping technologies, cloud services, and deep automotive expertise. Customers using these technologies can start their development process earlier and run thousands of tests in parallel, as opposed to running a much smaller number of tests sequentially on a limited number of physical testbenches.