The engineering function is – still – broken. Here is how to fix it.
Engineers spend roughly half their day on documentation, reporting and information search. The fix is not more AI tools. It is reinventing the engineering function as a system, built on a digital core and a single source of truth, the digital data thread.
This is the main finding of new Accenture research on “Reinventing for Human + AI Engineering”. The research provides five structural moves to shift engineering from cost center to growth engine.
Lets reflect what this means for the High Tech and Semicon industry
1. Run the V-model as a continuous evidence system | In lithography and precision equipment, one late requirement gap can cost months of re-verification. Requirements linked to architecture, tests and approvals in real time is what PLM and ALM systems need to actually deliver on.
2. Move to model-based, simulation-first development | Physical iteration in semiconductor equipment is expensive. The unlock is connecting MBSE models to simulation environments so architectural decisions propagate before anything gets built.
3. Automate verification and compliance at scale | A software update in semicon equipment can affect machine qualification at a customer fab. Evidence continuously linked across ALM, PLM and test management is not a nice to have, it is a competitive differentiator.
4. Redesign the talent model for AI augmented engineering | The priority is not just AI upskilling, it is capturing deep system knowledge before it walks out the door as demographics shift. Hybrid profiles across mechanical, software and systems domains are what complex mechatronic development requires.
5. Make partner collaboration structured, not scrambled | A significant share of a litho system’s complexity lives outside the OEM’s walls. Structured co-development with shared baselines and governed data environments is the difference between a supply chain that enables speed and one that introduces risk.
Summary | All five moves depend on the engineering data backbone first. Integrating PLM, ALM, simulation and test management across systems never designed to talk to each other is the foundation every other move is built on.
What is holding your engineering function back most, the data backbone, the toolchain, engineering function fragmentation or the ways of working?
🔗 https://www.accenture.com/us-en/insights/industrial/reinventing-human-ai-engineering?c=acn_glb_theengineeringrleader_14281239&n=smc_0526