Physical AI is changing how automotive engineering teams design, validate, and manufacture complex systems. However, delivering value requires more than training a model. In this fireside chat, specialists from CoreWeave’s Physical AI Field Engineering team— Isabel Ashworth, James Allibone, and Kacper Waniek—will explore how engineering teams can turn existing build and test data into practical, validated workflows to support better calibration decisions, reduce reliance on repetitive physical testing, and maintain confidence in the results. Through a candid discussion grounded in field experience, the speakers will examine what it takes to move from an engineering question and a data set to a workflow that domain experts can understand, evaluate, and use in practice. You’ll also hear perspectives on how domain engineering, machine learning, and specialist field engineering work together to move from proof of concept toward usable engineering tools.
What to expect
A practical look at the engineering challenges behind applying Physical AI to calibration and testing workflows
A discussion of the journey from data assessment and feature engineering through model development, benchmarking, and evaluation against an existing engineering process
Perspectives on how domain engineering, machine learning, and specialist field engineering work together to move from proof of concept toward usable engineering tools
What you’ll learn
How historical build and test data can be used to investigate whether calibration outcomes can be predicted before every physical test is run
Why data quality, physics-informed feature engineering, transparent evaluation, and explicit model limitations matter in safety- and quality-critical engineering workflows
What makes a model trustworthy to engineers: clear explanations of what it predicts, what data drives it, how performance is measured, where it fails, and how engineers can review the output
Why moving from proof of concept to adoption requires more than model accuracy, including usable interfaces, clear engineering success criteria, and a practical implementation path
How forward-deployed specialists help bridge the gap between AI capability and the engineers who need to use the resulting workflow in practice
Who should attend
Automotive engineering, R&D, manufacturing, and product development leaders
Calibration, validation, testing, controls, electrical, mechanical, and systems engineers
AI, machine learning, and data science leaders and practitioners working with engineering or manufacturing teams
Webinar
