Situation

A multimodal transport company with a fleet of trucks, forklifts and transshipment equipment spread across various terminals in the Netherlands and Belgium. Operational continuity is crucial for their services.

Challenge‍

Vehicles and equipment regularly went out unplanned, leading to chain disruptions, fines and extra staff deployment. Maintenance was carried out at fixed intervals or visual inspection, without taking into account actual use or wear.

Solution‍

A predictive maintenance model was developed based on real-time IoT sensor data (such as temperature, vibrations and frequency of use) in combination with historical maintenance logs. This enabled technical teams to intervene proactively before a failure occurred.

Approach

1. Inventory of equipment and sensors
   For each vehicle type, we identified which sensors were available and which data was relevant for maintenance forecasting.
2. Data collection and labeling
   Real-time sensor data was collected and labeled based on known failures in the past.
3. Model development
   We developed a machine learning model that recognized patterns prior to defects. Alerts were generated in case of abnormal behavior.
4. Integration with maintenance planning
   The system was linked to maintenance management, so that reports were immediately converted into maintenance tasks.

Results

• 30% fewer unexpected downtime
• €250,000 savings on maintenance costs in the first year
• Higher equipment reliability
• Fewer operational planning disruptions

Learnings

By making smart use of data from existing sensors, the company transformed its maintenance strategy from reactive to predictive. The investment paid off within a year, and the reliability of the operation increased significantly.

‍