How can PLM benefit from AI?
Within the wider IT landscape, PLM systems are not the only element in operation: PLM might typically represent a maximum of around 30% of the systems used in a company. However, in terms of data, we would estimate that around 70% of the strategic data needed to build a product is stored in PLM systems.
To date, however, that data has not been efficiently exploited, and the promise of PLM has not been fully realized. Here are common challenges companies face that AI can solve.
Companies often struggle with PLM complexities such as these:
Designing, implementing, and maintaining PLM software is a highly complex task.
A PLM system has to manage and store highly complex data and relationships (versions, configurations, maturities, etc.).
PLM involves many different disciplines, and so coordination and alignment are key.
Achieving end-to-end traceability along the value chain requires complex processing.
This traceability issue is just one aspect of process complexity. When creating end-to-end lifecycle services such as requirement validation, traceability, and change impact management, most companies have relied on point-to-point data connections between the bills of materials (BOMs) of different disciplines. This approach has been successful locally – notably in integrating the engineering BOM (EBOM) with the manufacturing BOM (MBOM).
Despite its potential, anchoring the entire suite of product lifecycle services solely on BOM-level connections has introduced significant operational challenges. The sheer volume of data, coupled with the intricacies of configuration management, has rendered this approach unwieldy and overly complex from a business standpoint.Companies therefore need ways to decouple the task of connecting the many kinds of BOM data from the provision of end-to-end lifecycle services. This means managing data in a way that allows work to be decentralized without any loss of integrity or consistency.
The figure contrasts the traditional way of working, with numerous links between the different BOMs at different stages of the lifecycle, with an innovative knowledge-based approach that enables data to be accessed wherever it is needed, regardless of where and how it is stored.
Parametric design and the closely associated activity of simulation form an integral part of generative product design and development. The new technologies provided by PLM should help, but until now the promise of end-to-end lifecycle optimization has not been completely realized. In particular, facilities have been too complex, and so adoption rates have been poor. Companies now need to act fast to modernize their process. Areas in urgent need of improvement include scalability, reconfigurability, and ecosystem integration.
