At the Share PLM Summit, one theme stood out above all others: artificial intelligence alone is not enough to transform engineering and manufacturing organizations. According to the Beyond PLM article, AI’s effectiveness depends on access to what many participants called “product memory,” the accumulated knowledge, decisions, relationships, and context surrounding a product throughout its lifecycle.
The author explains that many companies possess vast amounts of product data stored across PLM, ERP, CAD, and other enterprise systems. However, data by itself does not provide the context needed for meaningful decision-making. Product memory encompasses the reasoning behind design choices, trade-offs, assumptions, lessons learned, and organizational experience that often remain scattered across documents, emails, meetings, and the minds of experienced employees.
Several summit discussions focused on the growing excitement around generative AI and large language models. While these technologies can improve access to information and automate routine tasks, they are only as valuable as the knowledge they can retrieve and interpret. Without a structured way to capture and connect product memory, AI systems risk producing incomplete, inaccurate, or misleading results.
A recurring message from speakers was that product memory is not solely a technology challenge. Human participation remains essential. Engineers, designers, manufacturing experts, and other stakeholders create, interpret, and validate knowledge throughout a product’s development. Their expertise provides the context that transforms isolated data points into actionable insights. As a result, organizations must invest not only in digital platforms but also in processes and cultures that encourage knowledge sharing and preservation.
The summit also highlighted the limitations of traditional PLM implementations, which often focus on managing files and workflows rather than capturing the deeper context behind engineering decisions. Future systems, the author argues, should support richer knowledge networks that connect people, processes, and information.
The article concludes that the future of industrial AI will depend on a combination of advanced technology and human knowledge. AI may become a powerful tool for navigating complexity, but its value will ultimately be determined by the quality of the product memory organizations create, maintain, and share.