Portable Predictive Maintenance Device for Conveyor Rollers

Post by: Tobias Larsson
11th June 2026
No Comment

Preventing costly factory downtime and material waste through portable predictive maintenance.

We designed and built a portable predictive maintenance device for conveyor rollers to prevent costly unplanned downtime in manufacturing environments. Working with NKT HV Cables, our multi-modal sensor tool successfully detected early mechanical degradation by combining vibration, acoustic, and thermal data. This validated prototype offers a reliable, low-cost approach to reducing material waste and financial losses on the factory floor.

Programme: MSc Mechanical Engineering

Course: MSc Thesis

Semester: June 2026

Corporate partner: NKT HV Cables

The Challenge

Who has this problem?

Manufacturing facilities, specifically maintenance technicians at companies like NKT HV Cables, who rely on continuous cableway conveyor systems.

What is the consequence of the problem today?

Currently, roller maintenance is largely reactive or relies on subjective manual inspections. When a conveyor roller fails unexpectedly, it causes costly unplanned production downtime, severe material waste and energy loss.

What was your specific design challenge?

The challenge was to design and validate a portable, cost-effective predictive maintenance device capable of detecting early mechanical wear before a catastrophic breakdown occurs. The tool needed to fuse multiple data points (vibration, acoustic, and thermal) while remaining simple for factory floor workers to deploy without adding permanent, unnecessary complexity to the existing infrastructure.

The Solution

We designed and built a portable Condition Monitoring Device to provide low-cost predictive maintenance for conveyor systems. Instead of permanently installing expensive sensors, our device magnetically attaches to suspected problem areas for targeted monitoring.

The tool combines a 3-axis accelerometer, an infrared temperature sensor, and a digital acoustic sensor. By utilizing a "sensor fusion" strategy, cross-referencing vibration spikes with localized heat and noise, we effectively eliminated false alarms. We also integrated a local web server, allowing factory workers to easily monitor live data without requiring complex IT infrastructure.

Impact & Results

Our portable condition monitoring device provides a reliable early warning system for NKT HV Cables, directly reducing the risk of costly unplanned production stoppages. By identifying mechanical degradation before total failure, the solution cuts down on material waste and the unnecessary energy consumption.

During validation, our prototype successfully quantified roller health under operational stress. For example, under a 44.7 kg load, it registered a clear radial vibration peak of 1.41 m/s² for a failed roller, compared to just 0.51 m/s² for a healthy one. For factory operators, this means they can temporarily attach the device to a suspected problem area and receive objective, data-driven confirmation of its health, eliminating manual guesswork and improving overall system reliability.