Case Study: Vitesco Technologies achieves faster prototyping and reduced development time in powertrain control with MathWorks Reinforcement Learning Toolbox and Simulink

Vitesco Technologies Applies Deep Reinforcement Learning in Powertrain Control

Vitesco Technologies, a developer of vehicle electrification technologies, faced the challenge of building closed‑loop powertrain control that can handle a wide variety of environmental conditions while meeting tighter emissions rules and accelerated prototyping timelines. To meet these demands, Vitesco partnered with MathWorks, using Simulink along with Deep Learning Toolbox and Reinforcement Learning Toolbox to explore deep reinforcement learning approaches for powertrain control.

Using MathWorks tools, Vitesco engineers created a detailed plant model in Simulink and rapidly prototyped, generated, and optimized reinforcement learning agents with the Reinforcement Learning Toolbox and Deep Learning Toolbox. The MathWorks solution delivered fast prototyping and considerably reduced development time, enabled quick starts via documentation and examples, and sped problem resolution through dedicated MathWorks expert support.

Vitesco

Vivek Venkobarao

Vitesco