A smart framework for the availability and reliability assessment and management of accelerators technical facilities

CERN operates and maintains a large and complex technical infrastructure that serves the accelerator complex and experiments detectors. A performance assessment and enhancement framework based on data mining, artificial intelligence and machine-learning algorithms is under development with the objective of structuring, collecting and analysing the operation and failure data of the systems and equipment, to guide the identification and implementation of adequate corrective, preventive and consolidation interventions. The framework is designed to collect and structure the data and identify and analyse the associated driving events. It develops dynamically functional dependencies and logic trees, descriptive and predictive models to support operation and maintenance activities to improve the reliability and availability of the installations. To validate the performance of the framework and quality of the algorithms, several case studies are being carried out. In this paper, we report on the design and implementation of the performance assessment and enhancement framework, and on the preliminary results inferred on historical and live stream data from CERN's technical infrastructure. Proposals for the full deployment and expected long-term capabilities will also be discussed