We found that during package implementations, data collection and data modeling often gets less importance than other activities, which leads to data quality issues. Some of the reasons we found are:
• Unavailability of data - This typically happens in new implementations. Though companies would have basic data like Assets, Items etc but the data elements which are used in KPIs and decision making like Failure Hierarchy and Asset Attributes are missing.
• Data Loaded from multiple sources - Data exists in different files, applications and databases. Each data source is having its own format, granularity and completeness. Once this data gets loaded and consolidated into one system, there are discrepancies.
• Lack of Control - In many companies there is no proper control mechanism while defining new Assets, Items and Specifications etc. This leads to individuals defining the data based on their own standards and formats.
• Continued Manual Usage -Due to volume, compliance and complexity etc, sometimes companies continue using manual systems. For Example, safety related procedure. Though it is possible to have this implemented in package, but companies hesitate to implement them in system. This leads to suboptimal usage of system due to combination of manual process and software application.
This is just an indicative list and there are many more reasons like too much data, too less data, data granularity etc which also result into suboptimal usage of system. Typically these reasons lead to Excess Inventory, Low Asset Reliability, Non-Compliance to safety, Non-Compliance to standard maintenance procedures, reduced decision making capabilities and incorrect reporting.
While these reasons are the implementation reasons and a good implementation methodology would prevent these issues to happen, but at the same time these issues can crop up even after the implementation is over, due to a continued usage of the system and lack of control. Hence data cleansing and master data management is always a continuous activity and should not be considered just onetime activity.
To provide a complete data enrichment offering, some of the solutions our EAM and BPO practices are working on are:
• Tool based data profiling solution
• Data classification as per standards
• Data Cleansing
• Domain expertise for various industries
• Best practices for master data management
I will write more about our offering in next blog, once we have got our solution in some shape.