Are you ready for SAE Levels 4 and 5? Unlock highly automated driving systems verification - ERP Spain.com - España y América Latina

 

Before we can sleep tight into moving autonomous vehicles, there are some remaining gigantic challenges. The first challenge is the complexity and size required with verification to release highly automated driving (HAD) systems to the large public, which most automotive players have yet to overcome. While it’s now apparent simulation will soon become the predominant source of HAD verification, some questions still need answering:

 

- Which scenarios should be used for simulation?

- How to produce them in large numbers and high diversity?

- How to capture enough tricky adverse situations?

- When is the system good enough?

- When am I done testing?

- When is it considered complete?

 

 

The second challenge concerns data management and the need for extensive integration of application lifecycle management (ALM), product lifecycle management (PLM), and simulation tools. These tools can help keep track of ever-increasing complexity systems and conduct design, integration, and verification activities in harmonious collaboration between numerous departments.

 

 

Collaboration among an increasing number of contributors is necessary to advanced driving functions development. However, this increases the risk of information not being shared, synchronized, or traced across departments. Better coverage and increased verification confidence come with more simulation test cases, but it also means higher costs in computing infrastructure and workforce. Potentially more development insight comes with a more extensive result database, but it also adds more time to analyze the data without missing critical failure patterns.

 

In this webinar, you will learn how Simcenter Prescan360 can organically become part of autonomous vehicle development, providing the required verification by adequately combining:

 

- Extensive critical scenario generation according to SOTIF

- High fidelity sensor and environment simulation

- Large scale simulation on cloud and on-premises cluster

- Directed test case sampling using machine learning

- Meaningful representation of large results sets, not to miss significant failures

 

Enguerrand Prioux:ADAS/AD Product Line Manager, Siemens Digital Industries Software