Categorization Errors for Data Entry in Maintenance Work-Orders

In manufacturing, there is a significant push toward the digitizationof processes and decision making, by increasing thelevel of automation and networking via cyber-physical systems,and machine learning methods that can parse usefulpatterns from these complex architectures. As such, this pushtoward being “smart” is largely driven by the availability ofdata, for analysis, decision guidance, and the training of AI.The maintenance team, however, one of the core subsystemsin any production line, remains a largely human endeavor.Consequently, the historical data needed for research and developmentof AI-assisted maintenance frameworks are oftenfull of misspellings, jargon, and abbreviations. While onemight enforce data-entry into pre-specified functional categories(generally using some form of controlled vocabulary),cognitive models, along with consistent reports from industry,indicate that data entry remains a process fraught with significanterrors, especially when a mismatch occurs between designatedschemas and the technician’s needed semantic flexibility.This paper offers a framework for understanding andaddressing these issues, with a methodological case study inapplying Human Reliability Analysis (HRA) to quantify andunderstand human errors associated with entering maintenancework-order (MWO) data into structured database (DB)schema. We subsequently suggest potential mitigation strategiesfor each to improve the quality of recorded data throughoutthe maintenance-management workflow.