EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

50. Preventive maintenance for computerised tomography (CT) scan machines: an anomaly detection approach based on log-data.

Invited abstract in session MA-28: Advancements of OR-analytics in statistics, machine learning and data science 1, stream Advancements of OR-analytics in statistics, machine learning and data science.

Monday, 8:30-10:00
Room: 065 (building: 208)

Authors (first author is the speaker)

1. Felipe Maldonado
Mathematical Sciences, University of Essex

Abstract

Preventive maintenance is a very critical topic for many industries. Particularly in healthcare, often the associated costs (monetary and social) of unplanned downtimes are much higher than those coming from scheduled maintenance. However, deciding upon optimal schedules is a very complex task. Ideally, maintenance should be performed only when it is absolutely needed, minimising potential disruptions. In this paper we collaborate with Siemens to analyse a large data set consisting in log-data obtained from Computerised Tomography (CT) scan machines, with the objective of developing preventive maintenance based on anomaly detection. Here, the end-user (e.g., Hospitals) will have to make the decision of performing such maintenance based on the model predictions (and therefore, having to re-schedule patient's appointments). Having this into consideration we further develop explainable analytics that provide a visual representation containing warning time windows for future potential machine failures, and an explanation of why our model classifies them in that way.

Keywords

Status: accepted

Back to the list of papers