[Case Study] Streamlining the Evaluation of Pedestrian Protection Performance and Further Technical Advancement

Reduced performance prediction for each model from approximately 40 hours to 10 seconds! Performance evaluation possible without using CAE.

We would like to introduce a case where Honda R&D used our AI consulting services to streamline pedestrian protection performance evaluation and further deepen their technology. The company attempted to explore solutions from both pedestrian protection and rigidity perspectives, but felt there were limitations to this approach. Therefore, they tried a technique that involved machine learning to reproduce various conditions. As a result of the implementation, the performance prediction time, which previously took about 40 hours per model, was reduced to approximately 10 seconds. 【Case Study】 ■ Software Used: modeFRONTIER ■ Objective: To explore solutions from both pedestrian protection and rigidity perspectives ■ Challenge: There may or may not be common design variables in optimization calculations ■ Method: Constructing an agent that can evaluate performance without using CAE ■ Result: Reduced performance prediction time from about 40 hours per model to approximately 10 seconds *For more details, please refer to the PDF document or feel free to contact us.

simulatorOther analyses