Designing Intelligence for the Shop Floor: A Software Architect’s Guide to Industrial AI

This talk explores what it really takes to make "intelligent systems" work in the messy, high-stakes reality of production environments“ not just in demos or pilots. Most AI initiatives do not fail because the algorithms are weak, but because the surrounding system is not designed to handle uncertainty, change, and operational demands.

This talk offers a software architect’s perspective on designing intelligent, closed-loop AI architectures that operate reliably on the shop floor. Drawing on real industrial projects, we will highlight critical lessons learned: what worked, what failed, and why.

These insights cover technical, organizational, and architectural problems encountered in the plant.

We will share pragmatic design principles and proven architectural patterns for building end-to-end AI solutions, from sensors and programmable logic controllers (PLCs) to edge inference and local retraining. Attendees will leave with a clear mental model and a set of practical design ideas for turning clever prototypes into robust, understandable, and adaptable intelligent systems that people on the ground are willing to rely on.