TS TECH Integrates Python-Based AI Models with Simulation and Hardware Deployment
Engineers Successfully Developed a Real-Time System for Estimating Driver Fatigue
“We were able to realize a service that combines machine learning algorithms built in Python with other functions in real time using Simulink.”
Key Outcomes
- MATLAB and Simulink helped reduce the time required to integrate AI models with vehicle control functions, improving the development process and accelerating time to market
- MATLAB and Simulink enabled the design and testing of sophisticated driver assistance systems, leading to improved vehicle safety and enhanced driver comfort through accurate real-time fatigue estimation
- Embedded Coder ensured that the generated C code was optimized for real-time performance, resulting in a more reliable and efficient ECU deployment
TS TECH, a leading automotive interior systems supplier, faced the challenge of enhancing driver safety by integrating AI models for fatigue detection into traditional vehicle control models built in Simulink. This integration was crucial for enhancing vehicle interior comfort through advanced AI technology. To address this challenge, TS TECH used MATLAB® and Simulink® to bridge the gap between Python®-based AI models and ECU development.
MATLAB was used for developing and testing AI models and facilitating the conversion of Python-based AI models so that they can be simulated in Simulink. Simulink provided a platform for modeling, simulating, and testing the integrated system. This allowed the creation of complex simulation models that included both AI components and vehicle control functions. Additionally, Automated Driving Toolbox™ was used to design and test advanced driver assistance systems (ADAS), while optimized C code was generated with Embedded Coder® from the Simulink models for ECU deployment. With the support of MathWorks Consulting Services, TS TECH successfully developed a real-time system for estimating driver fatigue, showcasing the integration of AI with simulation and hardware deployment.
