)
)
)
How can we help you?
)
Marija Mijic
HR & Office Allrounder
T +43 660 383 95 46
office@craftworks.at
craftworks GmbH
Schottenfeldgasse 20/6a
1070 Vienna
Resource Efficiency with
Predictive Maintenance
Predicting gearbox failures for cost efficient processes and lower emissions.
:quality(80))
Success Story
reading time: 3 min
Automotive
Manufacturing
Predictive Maintenance
Visual Inspection
60%
Gearbox failures identified with AI
Predictive
Failures within the warranty period
Lower
Cost processes and emissions
Context
Due to a high number of truck failures, the truck manufacturer wanted to use predictive maintenance to detect component failures in the early manufacturing stages and repair them within the product's warranty period. One of the most serious failures are gearbox malfunctions.
Challenge
Gearbox failures are one of the most serious failures. By using an algorithm based on AI, it should be possible to detect transmission failure at an early stage. The data labeling was not completely accurate which caused a high level of class imbalance.
Assignment
The objective was to identify correlations and factors influencing gearbox damage, leading to the creation of a straightforward predictive model.
Solution
Correlations and influences on gearbox damage have been identified, leading to the creation of a predictive model utilizing XGboost to anticipate potential damage occurrences. The model was developed using Python's data science stack, incorporating NumPy, pandas, scikit-learn, and XGboost.
Through predictive maintenance and AI, this case exemplifies improved efficiency and higher margins by early detection of critical gearbox failures, showcasing the power of data-driven decision-making in operational enhancements.
:quality(80))