Human-Centred, AI-Enabled Production Scheduling

Our objective is to develop an AI-enabled scheduling framework that delivers adaptive, real-time, data-driven production scheduling under human and operational constraints. The approach combines:

• Multi-agent reinforcement learning for high-dimensional, dynamic scheduling optimisation (beyond throughput, e.g., energy/waste/carbon reduction).

• LLM/agentic AI to translate schedules into clear, actionable worker instructions (mobile/wearable), and to capture structured worker feedback in natural language.

• Explainability to make trade-offs understandable and increase trust.

• Plug-and-produce integration, leveraging generative AI to help configure and maintain interfaces to ERP/MES/SCADA.

We are looking for partners, especially:

• Application partners interested in advanced scheduling and/or currently facing scheduling challenges

• Industrial partners with a focus on production optimisation (e.g., scheduling, energy efficiency, waste reduction), preferably with experience or interest in machine learning, combinatorial optimisation, and/or LLMs/agentic AI

• System integrators / industrial IT/OT integrators

• Manufacturing companies willing to provide pilot lines, real constraints, and evaluation of usability/adoption.

• Worker-centric technology providers (wearables, HMIs, digital assistance systems).

• Social Science and Humanities Research Organizations