Data Quality Measurement

There is a saying in reliability, availability and maintainability (RAM) engineering with respect to data analysis: rubbish in, rubbish out. If you analyse poor quality data, you will obtain poor quality results or outputs. It is therefore very important to be able to measure the quality of your data to not only estimate the quality of your results, but so you can take active steps to improve your data quality.

The following is a small example of the types of data quality issues which can occur with maintenance and failure data. Data Quality Measurement seeks to determine the extent of these type of errors, and Key Performance Indicators can be created to drive data quality improvements in your organisation.

  • Ineffective recording of life units (hours, km, days) at maintenance events.
  • System age metrics showing negative or excessive aging due to input data errors.
  • Ineffective recording if the maintenance event was initiated in response to a failure or because the tradesmen predicted failure before the next scheduled maintenance event.
  • Maintenance actions being recorded in free text fields which provide limited ability for analysis
  • Work Orders which have start dates after their end dates,
  • Determining when equipment was ‘Out of Service’ from Work Order status codes,
  • Maintenance actions not being linked to the failure mode observed or prevented
  • Linking information from multiple tables within the CMMS.
  • Multiple work orders being open on the same equipment
  • Parts being ordered outside the work order so not recorded against the equipment
  • Inability to link deeper maintenance actions (engine rebuilds) to the initial failure
  • Obtaining consistent definitions and data entry compliance across multiple organisations (including maintenance conducted by external maintenance organisations)

Acuitas can assist you with measuring the quality of your data and developing validation rules improve data quality on entry to your system, or KPIs to change your data quality culture. Poor data quality does not necessarily ring the death knell for data analysis: Acuitas can assist you in ‘uncertain data analysis'.