Call for Participation
----------------------
We invite you to participate in our Predictive Maintenance Competition. This is the shared task of the XAI for Predictive Maintenance Tutorial at ECAI 2023 (
https://slawomirnowaczyk.wixsite.com/ecai-xpm).
Competition website:
https://slawomirnowaczyk.wixsite.com/ecai-xpm
Kaggle website:
https://www.kaggle.com/datasets/pattinson9999/uci-metropt-3-dataset
Important dates
----------------------
· 25/09/2023: submission deadline.
----------------------
Description
The compressed air system is a crucial component of the train and delivers essential pressurized air to several clients like:
· pneumatic suspension
· oil injection on the rail to reduce the friction and noise on the curves
· injection of sand to gain traction rails
· connect other trains
Requirements
Predict a failure or detect an anomaly two hours in advance (minimum time required by the company to schedule the replacement of the faulty train).
Description
· The dataset consists of more than 15 million examples
· Collected from February to August 2020
· 15 features from 7 analogue (1-7) and 8 digital (8-15) sensors
Failures
· Air leak: relevant sensors: tp2, tp3, h1, dv_pressure, comp, motor_current
· Oil leak: relevant sensors: oil_temp, oil_level
Possible Tasks:
· Anomaly Detection
· Anomaly Explanation
· Failure Prediction
· Remaining Useful Life (RUL)
Solution:
· Best Posters will be available on the XPM project website and presented online during the Tutorial
· Open repository with solution
· Send the Technical Poster to
bve...@fep.up.pt
Best,
Organizers