The system is designed to increase engine calibration efficiency, and to help manage the increasing complexity of powertrain design and development. Secondmind and Mazda are also extending a two-year research and development collaboration to focus on advanced hybrid and electric powertrain control systems, and strategic Case applications.
Mazda is initially using Secondmind to calibrate ECUs that control the company's next-generation Skyactiv engine technology and expects Secondmind advanced machine learning to more than double the efficiency of its conventional engine calibration process. Mazda has been at the forefront of model-based design (MBD) innovation for decades and has chosen Secondmind as its machine learning partner to help manage the mounting complexity resulting from tighter emissions regulations, increasing consumer demands and underlying pressure to achieve greater development process and environmental sustainability.
Backed by more than six years of practical machine learning research and development, Secondmind Active Learning for Powertrain offers state-of-the-art machine learning models based on noisy, high-dimensional data, and enables rapid, automated, and intelligent experimentation. Early indications are that Secondmind Active Learning for Powertrain can help car makers cut engine calibration time by up to 50%, reduce data acquisition and processing costs, and prototype materials use by up to 80% and 40% respectively. The potential impact to environmental and development sustainability is substantial as a result.
Gary Brotman, CEO of Secondmind, said: "Mazda is a leader in model-based design and a pioneer in the adoption and successful implementation of advanced machine learning. We're excited that Mazda has chosen us as a partner to help take their innovations in powertrain design and development to the next level.
Eiji Nakai, executive officer in charge of powertrain development and integrated control system development from Mazda, said: "Secondmind's unique active learning technology will enable us to automate the engine calibration process, and we expect to more than double efficiency in this area. In the future, we expect to extend the same innovation to more areas, such as controls of advanced Case technologies. We will utilise Secondmind machine learning technologies to help us evolve MBD and develop more efficiently. We are convinced that Secondmind technology will effectively solve the most complex optimisation problems faced by many companies."