Fleet Of Tools Research Articles

Application of machine learning for fleet-based condition monitoring of ball screw drives in machine tools

Ball screws are frequently used as drive elements in the feed axes of machine tools. The failure of ball screw drives is associated with high downtimes and costs for manufacturing companies, which harm competitiveness. Data-based monitoring approaches derive the ball screw condition based on sensor data in cases where no knowledge is available to derive a physical model-based approach. An essential criterion for selecting the condition assessment method is the availability of fault data. In the literature, fault patterns are often artificially created in an experimental test bench scenario. This paper presents ball screw drive monitoring approaches for machine tool fleets based on machine learning. First, the potentials of automated machine learning for supervised anomaly detection are investigated. It is shown that the AutoML tool Auto-Sklearn achieves a higher monitoring quality compared to literature approaches. However, fault data are often not available. Therefore, unified outlier scores are applied in a semi-supervised anomaly detection mode. The unified outlier score approach outperforms threshold-based approaches commonly used in industry. The considered data set originates from a machine tool fleet used in series production in the automotive industry collected over 8 months. Within the observation period, multiple ball screw failures are observed so that sensor data about the transient phases between normal and fault conditions is available.

Digital Fleet Management: A Scalable Cloud Framework based on Data-driven Prediction Models

A scalable Digital Fleet Management system can be leveraged by most organizations with high-volume high-value assets. In such scenarios, predictive analytics for tool health becomes central, as it enables decision-making in terms of planning, maintenance, end-of-life replacement, tool selection, etc. End-to-end solutions, spanning all the way from gathering live tool data to the visual representation of tool health, are certainly of major interest.
 Long-term fleet management can be accomplished through a consistent evaluation of the fleet performance profile. Predictive analysis can anticipate maintenance needs and resultant downtimes, and in turn it helps improve scheduling of procurement and distribution of the fleet.
 This paper focuses on managing a fleet of thousands of downhole tools based on tool health condition and other variables – a very common use case in Oil & Gas Services. An end-to-end automated scalable cloud framework is described in detail. This framework integrates failure prediction models for each single asset in the fleet of tools. Based on historical tool data, the models generate tool risk indices (one index per asset) which correlate to the probability of tool failure during near-future field jobs. These risk indices can be used for optimal asset-to-job mapping. They also help in de-risking field operations by identifying tools for overhaul or retirement. The proposed method integrates the tasks of: fetching data from 200,000+ tools, performing feature engineering, modeling via Machine Learning (ML) , and visualizing into a cloud pipeline. Framework scalability becomes a key requirement as fleet size increases or decreases over time to match market demands. The framework also allows for the addition of new ML models to the platform by citizen data scientists, who are not cloud experts. Finally, it is shown who this framework provides systematic steps for sustenance of such large cloud platform.