Search for a command to run...
Abstract For two decades, the central problem in industrial engineering has not been designing systems that work on paper. It has been closing the gap between what a system is designed to do and what can be reliably delivered in real operating conditions. This gap is visible across industrial practice. In this paper, it is examined through heat exchanger design, Stage-Gate development processes, and cross-functional project execution. The emergence of artificial intelligence does not eliminate this gap. It relocates it. The primary limitation shifts from execution capacity toward the definition of objectives, constraints, and decision frameworks. Similar transitions have occurred before. Fire decoupled human capability from what the body could do unaided. Electricity decoupled production from muscle. Artificial intelligence introduces the same shift in cognitive work. What can be engineered is no longer bounded by the time one engineer needs to think it through. This paper argues that the current stage corresponds to an installation phase, in which technical capabilities expand faster than their organisational implications are understood. The analysis draws on practical experience in industrial R&D and programme management, and uses topology optimisation in thermal engineering as a representative example of constraint-based, AI-assisted design. Within this context, the role of the industrial practitioner evolves. The traditional focus has been on correct execution, understood as the reliable management of established processes. That focus is progressively replaced by what this paper calls architecting intent: the precise formulation of goals, constraints, and operational boundaries within which autonomous systems operate. The move from execution to architecting intent extends beyond efficiency. It reshapes organisational dynamics. Internal friction in complex projects does not disappear. It is redistributed, and sometimes concealed, by AI systems. Accountability is also challenged. When decision-making processes are no longer fully interpretable in human terms, established notions of responsibility no longer hold. The central implication is this: the defining capability of industrial leadership is shifting from solving problems to defining them with precision. From a systems perspective, this represents a relocation of the primary contradiction. The constraint moves from execution toward problem formulation and system definition.