Our sister project COROB recently sent us a short set of questions as part of its Cluster Conversations interview series, which brings together projects funded under the AI, Data and Robotics Partnership. COROB — Cooperative robotics powered by AI and data for flexible production cells — develops a flexible, cooperative and intelligent multi-robotic solution that combines inspection, monitoring, control and AI techniques to make industrial processes more efficient and more flexible. Like PERKS, it is funded under Horizon Europe (GA No. 101120640).
The exchange gave us a welcome opportunity to step back from deliverables and pilots and explain, in plain terms, what PERKS is trying to change in industry. Here is a summary of what we told them.
What PERKS is about
PERKS builds a framework that brings together artificial intelligence, knowledge graphs, and human expertise to capture, manage and execute procedural knowledge in dynamic industrial environments. Two components sit at the centre of that framework: the Procedural Knowledge Management System (PKMS) and the Procedural Knowledge Ontology (PKO). Together they allow robots, AI agents and human operators to share and reuse operational know-how across very different manufacturing and service settings.
The motivation is not abstract. European industry is dealing with workforce shortages, growing production complexity and a need for automation that can be reconfigured quickly. PERKS responds by turning static operating procedures into adaptive, machine-interpretable knowledge asset, and by keeping human expertise at the centre rather than at the margins.
Contributing to integrated AI, data and robotics value chains
COROB asked how PERKS contributes to the call’s ambition of optimising production and service value chains. Our answer: by refusing to treat data analytics, AI and robotics as three separate technology stacks. PERKS connects them through a shared semantic representation of industrial procedures, so that knowledge created in one context can be reused, adapted and executed by another system, another operator or another production line.
The added value is demonstrated across our industrial use cases ( manufacturing, CNC machine parameterisation, and microgrid optimisation), where we target faster and more reliable procedure execution, less downtime thanks to AI-assisted troubleshooting, better knowledge transfer from experienced to less experienced workers, and more flexibility when production requirements change.
Adaptability and human-centred innovation
The third question concerned adaptability and human-centred design. Industrial reality is moving targets: evolving requirements, changing equipment configurations, and operators with very different levels of experience. PERKS addresses this by combining procedural knowledge, contextual information and user feedback so that workflows can adapt in real time.
Practically, that means human-centred interfaces, conversational assistants and intelligent guidance systems that accompany operators through complex tasks while leaving oversight and decision-making authority with the human. Through the PKMS, procedures can be modified, personalised and reused according to the operational context and the capabilities of the person carrying them out. Pairing symbolic reasoning with learning-based methods lets robots and AI agents collaborate with human experts, so that automation complements skills instead of replacing them.
Thanks to COROB
We would like to thank the COROB team for the invitation and for the chance to compare notes between two projects working on adjacent problems.
________
PERKS is funded by the European Union’s Horizon Europe research and innovation programme under Grant Agreement No. 101120323. Views and opinions expressed are those of the authors only and do not necessarily reflect those of the European Union.