Reza Pedrami is CS – Group Canada senior system safety expert and has extensive Control System, Control Software and Functional safety experience. He has 15 years of experience in design, verification, and validation of embedded control systems for safety critical applications in the aerospace industry. His involvement in the system risk analysis and safety assessment of novel software products at the aircraft level gives him an excellent opportunity to understand regulation and certification aspect of the integration of the innovative products. His effort in integrating novel software solutions into the domain of safety critical applications leads to several patents in the areas of electronic control of the gas turbine engine. He has a master’s degree in mechanical engineering specialized in dynamic systems and control with several years of research experience in advanced control system.
Practical Guidelines and Methods to Verify and Validate AI/ML-Based System for Safety Critical Application
A systematic method to formulate the best practices in AI/ML technology is studied. This methodology is based on categorizing the best practices and standard guidelines in three-view model: Data Science, AI/ML pipeline, and Systems Engineering. A feasibility study is successfully performed to demonstrate how AI/ML best practices in three-view model addressing the objectives provisioned for the certification of AI/ML based system for safety critical applications. Additionally, standard checklist for each model view is developed as a practical guideline to verify and validate AI/ML based system.