Process Control with Highly Left Censored Data
Javier Neira Rueda, Andres Carrion Garcia

TL;DR
This paper addresses the challenge of monitoring industrial processes with highly left censored data, proposing new statistical methods and control charts to accurately assess process parameters despite extensive data censorship.
Contribution
It introduces a novel estimation technique and control chart specifically designed for processes with over 70% left censored data, improving process monitoring accuracy.
Findings
Effective estimation of process parameters with high censorship levels
Development of a new control chart for left censored data
Demonstrated improved process detection in simulations
Abstract
The need to monitor industrial processes, detecting changes in process parameters in order to promptly correct problems that may arise, generates a particular area of interest. This is particularly critical and complex when the measured value falls below the sensitivity limits of the measuring system or below detection limits, causing much of their observations are incomplete. Such observations to be called incomplete observations or left censored data. With a high level of censorship, for example greater than 70%, the application of traditional methods for monitoring processes is not appropriate. It is required to use appropriate data analysis statistical techniques, to assess the actual state of the process at any time. This paper proposes a way to estimate process parameters in such cases and presents the corresponding control chart, from an algorithm that is also presented.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdvanced Statistical Process Monitoring · Fault Detection and Control Systems
