The use of Digital Twin for predictive maintenance in manufacturing

TLDR

The study proposes a Digital Twin–based methodology to compute the Remaining Useful Life of manufacturing equipment for predictive maintenance. It models equipment in a digital twin, synchronizes the model with sensor data, and uses the simulation output to estimate condition and RUL. The approach successfully monitors and predicts machine condition non‑invasively, as demonstrated by a case study predicting an industrial robot’s RUL.

Abstract

This paper presents a methodology to calculate the Remaining Useful Life (RUL) of machinery equipment by utilising physics-based simulation models and Digital Twin concept, in order to enable predictive maintenance for manufacturing resources using Prognostics and health management (PHM) techniques. The resources and the properties of them are first modelled in a digital environment able to simulate the real machine's behaviour. Data are gathered by machines' controllers and external sensors to be used for the synchronous tuning of the digital models and their simulation. The outcome of the simulation is then used to assess the resource's condition and to calculate RUL. In this way, the condition and the status of the machines can be monitored and predicted as a result from the simulation of physics-based models, without invasive techniques of common predictive maintenance solutions. A case study is presented in this paper where the proposed methodology is validated by predicting the RUL of an industrial robot.