Every factory, every office, every service operation runs on flow. When flow breaks down (bottlenecks pile up, lead times stretch, productivity erodes…) managers reach for the tools they know: Lean, process mining, structured problem-solving frameworks. These paradigms have delivered enormous value over decades of industrial history. But they were built for a world without large language models, generative AI and intelligent automation embedded directly into ERP and MES systems. That world no longer exists.

AI is reshaping operations management faster than the discipline can keep up. The potential is obvious; smarter scheduling, real-time bottleneck detection, predictive process optimization, autonomous decision support. But the honest answer to the question “how do you actually integrate AI into your operations flow?” is that nobody has fully figured it out yet. There is no methodology. There is no playbook. The debate is wide open.

This thesis steps into that gap. Rather than starting from scratch, the research takes seriously what Lean Management, TSEF and Process Mining got right – the accumulated wisdom of decades of flow optimization – and asks how those foundations can be extended, not replaced, by AI. The output is a new methodology: a structured, evidence-based approach that connects the digital systems already running inside companies’ ERP and MES infrastructure with the capabilities that modern AI platforms now make possible.

The research path combines literature review, real-world case studies and hands-on experimentation and simulation, because a methodology that cannot survive contact with a real factory floor is not a methodology worth building.

If you want to work at the frontier where industrial engineering meets artificial intelligence and leave behind something that companies can actually use, this is the thesis that puts you there.