The speed of data creation is amazing these days. According to the last IBM research, 90% of the data in the world today has been created in the last two years alone. I’m not sure if IBM counting all enterprise data, but it doesn’t change much- we have lots of data. In manufacturing company data is created inside of the company as well as outside. Design information, catalogs, manufacturing data, business process data, information from supply chain – this is only beginning. Nowadays we speak about information made by customers as well as machined (so called Internet of Things).
One of the critical problems for product lifecycle management was always how to feed PLM system with the right data. To have right data is important – this is a fundamental thing when you implement any enterprise system. In the past I’ve been posted about PLM legacy data and importance of data cleanup.
I’ve been reading The PLM State: Getting PLM Fit article over the weekend. The following passage caught my special attention since it speaks exactly about the problem of getting right data in PLM system.
[...] if your data is bad there is not much you can do to fix your software. The author suggested focusing on fixing the data first and then worrying about the configurations of the PLM. [...] today’s world viewing the PLM as a substitute for a filing cabinet is a path to lost productivity. Linear process is no longer a competitive way to do business and in order to concurrently develop products, all information needs to be digital and it needs to be managed in PLM. [...] Companies are no longer just collecting data and vaulting it. They are designing systems to get the right data. What this means on a practical level is that they are designing their PLM systems to enforce standards for data collection that ensure the right meta data is attached and that meaningful reports can be generated from this information.
PLM implementations are facing two critical problems: 1/ how to process large amount of structured and unstructured information prior to PLM implementation; 2/ how constantly curate data in PLM system to bring right data to people at the right time. So, it made me think about importance of data curation. Initially, data curation term was used mostly by librarian and researchers in the context of classification and organization of scientific data for future reuse. The growing amount and complexity of data in the enterprise, can raise the value of digital data curation for implementation and maintenance of enterprise information systems. PLM is a very good example here. Data must be curated before get into PLM system. In addition to that, data produced by PLM system must be curated for future re-use and decision making.
What is my conclusion? The complexity of PLM solutions is growing. Existing data is messy and requires special curation and aggregation in order to be used for decision and process management. The potential problem of PLM solution is to be focused on a very narrow scope of new information in design an engineering. Lots of historical record as well as additional information are either lost or disconnected from PLM solutions. In my view, solving these problems can change the quality of PLM implementations and bring additional value to customers. Just my thoughts…