Sensor Monitoring Towards Industry 4.0: Bridging Concepts on the $C_n$ Multivariate Control Chart

The main challenge in monitoring multiple sensors is the high dimensionality that a process can easily reach. From a decision-making perspective, converting these multivariate dynamics into an informative visualization is also not trivial. Statistical Process Control (SPC) techniques detect changes in quality characteristics, but they are difficult to apply beyond the normality (Gaussian) assumption. For instance, the most well-known SPC multivariate control chart is based on Hotelling's $T^2$ statistic, which assumes that the process data follow a multivariate normal distribution (which rarely occurs in practice) and that the observations are serially independent. This study focuses on the multivariate case (dimension $d \geq 2$), regardless of the data dependence or marginal asymmetry. We propose a new control chart for change detection in copula-based multivariate processes, called the $C_n$-chart, which is based on a density level-set tolerance region built from a copula function, and we derive its run-length properties. The methodology also quantifies the (not necessarily linear) dependence among the variables. Two case studies are presented, one on water quality control and the other on monitoring brass-steel thermostats, complemented by a simulation study of run-length performance and computational cost. The results show that the proposed methodology captures both the association among variables and the anomalies, and that it accommodates variable selection through conditional (regression) modeling of the margins.

Authors

Publication Details

Journal
arXiv (Cornell University)
Published
2026-11-18
DOI
https://doi.org/10.48550/arxiv.2611.23820
Primary Topic
Advanced Statistical Process Monitoring
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Sensor Monitoring Towards Industry 4.0: Bridging Concepts on the $C_n$ Multivariate Control Chart

Diego Carvalho do Nascimento, Francisco Louzada, Paulo H. Ferreira, Ana C. S. Batista
arXiv (Cornell University)
Advanced Statistical Process Monitoring
preprint

Sensor Monitoring Towards Industry 4.0: Bridging Concepts on the $C_n$ Multivariate Control Chart

Diego Carvalho do Nascimento, Francisco Louzada, Paulo H. Ferreira, Ana C. S. Batista
preprint en

Abstract

The main challenge in monitoring multiple sensors is the high dimensionality that a process can easily reach. From a decision-making perspective, converting these multivariate dynamics into an informative visualization is also not trivial. Statistical Process Control (SPC) techniques detect changes in quality characteristics, but they are difficult to apply beyond the normality (Gaussian) assumption. For instance, the most well-known SPC multivariate control chart is based on Hotelling's $T^2$ statistic, which assumes that the process data follow a multivariate normal distribution (which rarely occurs in practice) and that the observations are serially independent. This study focuses on the multivariate case (dimension $d \geq 2$), regardless of the data dependence or marginal asymmetry. We propose a new control chart for change detection in copula-based multivariate processes, called the $C_n$-chart, which is based on a density level-set tolerance region built from a copula function, and we derive its run-length properties. The methodology also quantifies the (not necessarily linear) dependence among the variables. Two case studies are presented, one on water quality control and the other on monitoring brass-steel thermostats, complemented by a simulation study of run-length performance and computational cost. The results show that the proposed methodology captures both the association among variables and the anomalies, and that it accommodates variable selection through conditional (regression) modeling of the margins.

arXiv (Cornell University)
Advanced Statistical Process Monitoring
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon