title: "Multi-State Adaptive-Dynamic Process Monitoring" author: Gabriel Odom, Ben Barnard, and Melissa Innerst

Overview

We create this package, mvMonitoring, from the foundation laid by Kazor et al (2016). This package is designed to make simulation of multi-state multivariate process monitoring statistics easy and straightforward, as well as streamlining the online process monitoring component.

Installation from CRAN

Install the stable version of this package via

install.packages("mvMonitoring")

Installation of Development Version

Make sure you have the latest version of the devtools package, and pull the package from GitHub.

devtools::install_github("gabrielodom/mvMonitoring")

Load the library after installation by

library(mvMonitoring)

Examples

These are the examples shown in the help files for the mspProcessData(), mspTrain(), mspMonitor(), and mspWarning() functions.

# Generate one week's worth of normal operating (NOC) data recorded at the one-
# minute level
nrml <- mspProcessData(faults = "NOC")
# The state values are recorded in the first column.
n <- nrow(nrml)

# Calculate the training summary, but save five observations for monitoring.
# This function will treat the first 3 days as in control (IC), and then update
# the training window each day.
trainResults_ls <- mspTrain(
  data = nrml[1:(n - 5), -1],
  labelVector = nrml[1:(n - 5), 1],
  trainObs = 4320
)

# While training, we included 1 lag (the default), so we will also lag the
# observations we will test.
testObs <- nrml[(n - 6):n, -1]
testObs <- xts:::lag.xts(testObs, 0:1)
testObs <- testObs[-1,]
testObs <- cbind(nrml[(n - 5):n, 1], testObs)

# Run the monitoring function.
dataAndFlags <- mspMonitor(
  observations = testObs[, -1],
  labelVector = testObs[, 1],
  trainingSummary = trainResults_ls$TrainingSpecs
)

# Alarm check the last row of the matrix returned by the mspMonitor function
mspWarning(dataAndFlags)

Paper Graphics

The R code to build and save the simulation graphics from the paper are in the inst/mspGraphsGrid.R file.

Acknowledgements

This work is supported by the King Abdullah University of Science and Technology (KAUST) Office of Sponsored Research (OSR) under Award No: OSR-2015-CRG4-2582 and by the National Science Foundation PFI:BIC Award No: 1632227.