As automotive OEMs face increasingly stringent emissions regulations, they must prioritize powertrain efficiency, whilst optimizing driveability and fuel economy to deliver high-performance vehicles that meet the needs of their customers.

Effective and efficient engine calibration is the key to balancing these competing objectives but requires meticulous control of numerous combustion parameters. Traditional model-based calibration processes can now no longer cope with the high-dimensional optimization challenge of modern engines, driving automotive manufacturers to explore advanced machine learning techniques to help solve this problem.

Secondmind for Calibration accelerates model-based calibration processes. It is powered by Secondmind Active Learning, an advanced machine learning system that intelligently automates the Design of Experiments (DoE), intuitively selecting only the most important data required to meet the optimization objective of the engineer.

This results in a data-efficient approach that has helped automotive OEMs achieve more accurate calibrations while cutting time spent on test benches by 50% and data requirements by 80% - enabling faster development cycles, lower costs and a shorter time to market.

This white paper explores:

-       The evolving landscape of automotive engine calibration 

-       Limitations of traditional model-based calibration methods

-       The benefits of applying advanced machine learning techniques to engine calibration processes

-       How Secondmind for Calibration works

-       Comprehensive case study of how Mazda Motor Corporation achieved a 59.7% reduction in human hours for model-based calibration with Secondmind

Download the full white paper here