A new proposal for replacing legacy product lifecycle management systems may repeat many of the problems it seeks to solve. Beyond PLM’s Oleg Shilovitsky examines the Future Platform for Hardware Engineering, or FPHE, proposed by Doug Macdonald as a successor to traditional PLM.
Shilovitsky agrees with FPHE’s diagnosis of a persistent engineering problem. Despite decades of PLM development, engineers still rely heavily on spreadsheets, PDFs, email, and collaboration tools to exchange information across disciplines. Artificial intelligence cannot solve this fragmentation simply by processing documents. AI agents need structured information about products, relationships, configurations, provenance, and access rights.
The disagreement concerns FPHE’s proposed solution. Shilovitsky argues that it repeats familiar PLM ideas under a new name, including broader lifecycle coverage, centralized data models, and promises of a single source of truth. A universal ontology is particularly problematic because engineering terminology and processes vary between organizations. Making the model flexible enough for individual companies can recreate the customization and implementation complexity associated with traditional PLM.
He also challenges the claim that established PLM systems remain centered primarily on mechanical CAD. Major vendors have expanded into requirements management, systems engineering, simulation, manufacturing, service, and other areas. The larger problem is that capabilities acquired over time often remain separate products with different data models and implementations, preventing customers from experiencing them as one coherent environment.
Shilovitsky sees more promise in emerging platforms that connect existing CAD, PLM, ERP, MES, and engineering systems rather than replace them.
The deeper opportunity, he argues, is product memory. Traditional systems record what a product is, while digital threads establish connections among formal records. Product memory would also preserve why decisions were made, including alternatives, trade-offs, changes, and engineering rationale.
Capturing this context could ultimately be more valuable than expanding PLM’s scope, particularly as AI agents become more involved in engineering decisions.