When failures hide in plain sight
In many industrial environments, predicting equipment failures is far from straightforward. Harsh operating conditions, limited accessibility, or simply the high cost of sensors make continuous monitoring impractical. Manufacturers like SEW Usocome, a leading motor producer and MODAPTO pilot partner, know this challenge all too well. With heterogeneous equipment in production cells and tight production schedules, they often face the reality that traditional condition-based maintenance (CBM) is simply not feasible.
At the 13th IMA International Conference on Modelling in Industrial Maintenance and Reliability (MIMAR 2025), this challenge was addressed through new MODAPTO research inspired by SEW’s operations. The work was presented by Hanser Jiménez, researcher at the Centre de Recherche en Automatique de Nancy (CRAN), Université de Lorraine, together with Phuc Do, Alexandre Voisin, and Chiara Franciosi, also from CRAN. Their joint contribution showed how smart hybrid policies can unlock cost savings and improve resilience—even when degradation cannot be directly observed.
MIMAR is widely regarded as the premier international forum for maintenance and reliability modelling, bringing together leading researchers and practitioners to discuss state-of-the-art advances in maintenance theory, predictive and reliability models, and industrial applications.
A new way to think about maintenance
The research introduces a threshold-based hybrid maintenance strategy. Instead of relying on sensors or fixed schedules, this approach looks at the sequence of system stops and failures. Each stop in SEW operations triggers an imperfect repair—a quick intervention that restores the system at low cost, but not fully “as good as new.” Over time, however, repeated repairs become less effective as assembly and disassembly operations add complexity and accelerate wear.
The innovation lies in setting a repair threshold: after a predefined number of repairs, the system is no longer patched up but instead receives a preventive replacement, which also respects the spare parts replenishment delay. This balances short-term savings with long-term reliability, avoiding the pitfalls of pure corrective or purely time-based strategies.
Why it matters for manufacturers like SEW Usocome
For SEW Usocome and similar manufacturers, this approach brings three key advantages:
- No dependence on sensors: Decisions rely only on failure history, not real-time monitoring, making it suitable for environments where CBM is unworkable.
- Adaptability to uncertainty: The policy dynamically responds to how often failures occur, which is crucial when degradation patterns are unpredictable.
- Reduced downtime and costs: By striking the right balance between quick, low-cost fixes and well-timed replacements, the strategy helps manufacturers cut maintenance expenses, avoid costly unplanned stoppages, and keep production lines running smoothly.
All in all, these advantages confirm its potential for real-world applications where failures are uncertain and spare parts logistics must be carefully managed.
Beyond theory: towards industrial impact
What makes this research particularly relevant for MODAPTO is its grounding in industrial reality. At SEW Usocome, maintenance teams often juggle operator variability, limited inspection opportunities, and the constant need to keep production flowing. A threshold-based strategy provides a practical decision rule: repair up to a point, then replace—simple to implement, yet powerful in reducing both risk and cost.
Moreover, by incorporating a replenishment delay, the model ensures that spare parts can be delivered in time for planned replacements. This feature respects supply chain constraints while avoiding long service interruptions—a common pain point in many factories.
Looking ahead
The next steps for MODAPTO’s research include extending this approach to a multi-component level, such as for SEW cell modules, where interactions between parts add complexity. The expectation is clear: in contexts where traditional models struggle due to uncertainty, threshold-based policies can become a robust alternative.
By aligning advanced modelling with the real challenges faced by manufacturers like SEW Usocome, this work showcases how MODAPTO is helping industry partners bridge the gap between research and practice—supporting digital transformation and smarter maintenance strategies across Europe.