The Wonderland of Operating the ALICE Experiment

ALICE is one of the experiments at the Large Hadron Collider (LHC), CERN, Geneva, Switzerland. Composed of 18 sub-detectors each with numerous subsystems that need to be controlled and operated in a safe and efficient way. The Detector Control System (DCS) is the key to this and has been used by detector experts with success during the commissioning of the individual detectors. During the transition from commissioning to operation, more and more tasks were transferred from detector experts to central operators. By the end of the 2010 datataking campaign, the ALICE experiment was run by a small crew of central operators, with only a single controls operator. The transition from expert to non-expert operation constituted a real challenge in terms of tools, documentation and training. A relatively high turnover and diversity in the operator crew that is specific to the HEP experiment environment (as opposed to the more stable operation crews for accelerators) made this challenge even bigger. This paper describes the original architectural choices that were made and the key components that enabled the DCS to come to an homogeneous control system that would allow for efficient centralized operation. Challenges and specific constraints that apply to the operation of a large complex experiment are described. Emphasis will be put on the tools and procedures that were implemented to allow the transition from local detector expert operation during commissioning and early operation, to efficient centralized operation by a small operator crew not necessarily consisting of experts.