Theme 04Registration open

AI in the quality system: from inspection to prediction, and what the standard still requires

Thursday, November 19, 2026 · 13:00 UTC

45 minutes plus questionsFree, online
Arjun MenonSofia MartinsDaniel WeberWith Arjun Menon, Sofia Martins and Daniel Weber
Other time zones
Thursday, 13:00 UTC18:30 India17:00 Gulf14:00 Central Europe13:00 London08:00 New York10:00 Sao Paulo21:00 Singapore, Manila

About this session

Machine vision now catches defects a line inspector cannot, anomaly detection flags a drifting process before the control chart does, and generative tools draft procedures, root-cause reports and audit responses in minutes. This session looks at what that has changed in the quality system of a plant and of a service center, and what it has not: the standard still requires a defined process, a competent person accountable for the decision, and evidence that the control worked.

Two cases, one from manufacturing and one from shared services, show where AI moved a quality result and where it created a new risk the audit found first.

What you take away

  • Which quality tasks AI has genuinely changed, with the measured effect
  • What a quality standard still demands from a human, and how to document it
  • A checklist for introducing an AI-assisted control without losing the audit trail

Who it is for

Quality managers, plant heads and service-center leads, and anyone preparing for a surveillance audit with AI tools in the process.

Speakers

Arjun Menon
Arjun Menon Head of Quality, Sundaram Precision Parts

Arjun leads quality across four machining plants. Over the last three years his team moved from end-of-line inspection to in-process prediction, cutting customer complaints by 38 percent while keeping IATF 16949 certification intact. He previously ran supplier quality for a Tier 1 automotive group and teaches the statistical process control module in the company academy.

Sofia Martins
Sofia Martins Data Science Lead, Kestrel Automation

Sofia builds the machine vision and anomaly detection models behind Kestrel's inspection systems, and works with quality engineers to decide which alerts deserve a stop.

Daniel Weber
Daniel WeberPlant Manager, Rheinfeld Components

Format

One speaker, one real case, forty-five minutes, then fifteen minutes for your questions. Sessions are free and open to everyone; recordings and materials go to members, and membership is free.

More from Kestrel Automation

More webinars

All webinars →

Go to Top