Asset performance data quality is vital for the success of your APM project

Resolving asset performance data quality issues with transactional data governance (TDG) is foundational to achieve results from your APM solution

There’s a common problem that exists across many SAP Maintenance and Service users – a lack of asset performance data quality and integrity. Asset performance data includes, amongst others, data about how the asset failed, why it failed, how long it was failed and what was done to repair it. Asset performance data is captured transactionally in Notifications using the Catalog codes during maintenance processing.

The objective of collecting and analyzing equipment performance data in APM is to make better equipment decisions and optimize plant operations. Yet high-quality equipment performance data is often unavailable due to pervasive data quality issues.

Before a company can reap the benefits by APM, and its AI and Machine Learning features, they must have a clear transactional data taxonomy and high fidelity in their asset performance data.

Many companies cannot generate meaningful performance metrics for even individual equipment items and the problem gets worse when trying to analyze asset reliability on plant units, systems and other structured asset relationships. The inability to perform accurate reliability analysis compromises asset management decisions, increases risk, and decreases profitability.

In attempts to get around poor performance data, companies turn to one or more of the following approaches to try to improve their asset management decision-making:

1. Data augmentation via AI and NLP
   AI and NLP can fill in some missing data but do not fix the root cause of data quality issues. They address the symptoms of poor-quality data versus the root causes.
2. Industry databases, e.g., CCPS PERD and the OREDA JIP
   Industry reliability data sources compile data from a small number of participating companies. However, participating companies have the same data quality issues as those companies seeking to use industry data. They qualify and manipulate their data at the end of 2-3 year data collection phases. Similar to data augmentation, periodic reliability data qualification addresses symptoms versus root causes and does not fix underlying performance data quality issues.
3. Master Data Governance (MDG) for Enterprise Asset Management (EAM)
   MDG for EAM addresses master data, not transactional data.5 Reliability data captured in Notifications is transactional; it records equipment failure events, preventive maintenance and inspection results, and risk assessment data.

These approaches can be useful but by themselves cannot achieve the goal of high-quality transactional performance data in a standard, normalized, complete, coherent, structured, timely, and accessible way. In short, there is a need for Transactional Data Governance of performance data.

What is Transactional Data Governance (TDG)?

Transactional data are the basis for the most important decisions for equipment in operating and producing facilities. Without these data we cannot calculate failure rate, analyze consequences or measure how failures impact our business objectives. Timeliness of performance data capture is key; if details aren’t collected at the time of an event, it is difficult to reconstruct details at a later point in time. This means data collection must be intuitive and accessible, preferably on a mobile device in the hands of the operators and maintainers. Performance data is also automatically recorded in APM based on rules. Transactional data governance is time sensitive.

In contrast, asset master data is usually not as time sensitive as asset master data is readily available from asset documentation, project books, purchase orders, or even observations from field walk-downs.

TDG can be used to prioritize Master Data Governance (MDG) efforts so focus can be given to equipment items with the highest consequence of failure, or rate of failure, and which most affect corporate objectives.

ISO 14224

Helpfully, the ISO 14224:2016 standard differentiates between transactional and master data, with specifications for equipment performance data in Table 3 – Reliability and maintenance parameters in relation to taxonomy levels, Table 6 – Failure data, and Table 8 – Maintenance data, and specifications for equipment master data in Table 5 – Equipment data common to all equipment classes, with reference to individual equipment data in Annex A. Figure 6 – Maintenance categorization, can be used to align data collection forms with data requirements, with each maintenance category representing one dataset and unique set of data requirements.

Transactional Data Governance (TDG) includes quality assurance (QA) and quality control (QC) steps. There are two parts to performance and failure data QA, (1) specification of data requirements and system validating compliance on user input, and (2) performing cause analyses of data quality issues to identify problem areas and put steps in place to resolve them.

For example, selecting incorrect reference objects for failure reporting is a pervasive data quality issue, one that can be resolved with the taxonomic classification of technical objects, and application of system validations automatically ensure only the correct objects can be selected.

Quality control is the partial sampling of completed records to ensure compliance with data specifications, with feedback, training, or realignment given to field personnel as required to remedy compliance issues.

Conclusion

Applying Transactional Data Governance improves performance data quality for all equipment assets and can be started right away, and especially before embarking on an APM initiative.

TDG pays back with high-quality (data-driven) decision-making almost immediately. A TDG initiative can be done in conjunction with am MDG initiative and they are complimentary.

Table 1. Transactional versus master data

References

1. Ciliberti, VA, 2021, ISO Course, data quality excerpt, “Use of ISO 14224 for Optimizing Safety and Profitability in the Oil and Gas Industry – in a Digitalized Perspective”
2. Markets and Markets, “Substantial gap between customer expectations and solutions delivered,” Enterprise Asset Management Market by Application, Component, Organization Size, Deployment Model, Vertical and Region – Global Forecast to 2026
3. Nagle, et al, 2017, “Only 3% of Companies’ Data Meets Basic Quality Standards,” Harvard Business Review.
4. Data Management Body Of Knowledge, Second Edition, 2017, DAMA International
5. Utopia MDG, MDG for EAM Web page