Targeted automation offers an Industry 5.0 route to smarter operations
Industry 5.0 thinking places value on technology that complements people rather than simply replacing them. A new research framework applies that principle to operational decision-making: automate cases where human intervention adds little, while preserving human input where it can improve both the immediate decision and the underlying AI.
The work was developed by Kai Hoberg of Kühne Logistics University, Christina Imdahl of Eindhoven University of Technology and William Schmidt of Emory University. Their focus is not the broad contest of human versus machine. Instead, they treat both as parts of a changing system in which automation handles repetitive work and employees provide targeted expertise.
From fixed rules to a learning feedback loop
The proposed process begins by identifying situations in which people improve algorithmic recommendations. A predictive model then estimates the likely value of intervention in future cases. Low-value interventions can be automated, while cases with higher potential benefit remain with human decision-makers.
Human corrections are subsequently returned to the algorithm. This feedback loop is important for industrial resilience because operating conditions do not remain constant. A system that continues to learn from expert input is better positioned to adapt than one designed only to maximise short-term automation.
High automation without removing expertise
The researchers studied more than 475,000 order-related decisions recorded over four years at a major commercial vehicle manufacturer with annual revenue above €4 billion. In the empirical application, around 84% of decisions could be automated.
That figure should not be read as a case for eliminating human involvement. It demonstrates the value of directing scarce expertise towards decisions where it has a measurable effect. For manufacturers, this can reduce repetitive workload, support more meaningful use of employee knowledge and retain a mechanism for improving AI as products, demand and processes change.
Companies considering this model should begin by measuring the added value of human intervention rather than setting an automation percentage first. That evidence can support more robust implementation decisions and a clearer balance between productivity, adaptability and human control. The research includes Kühne Logistics University.
