How Internet has changed the control hierarchy Mesarovic, Macko and Takahara published their seminal book “Theory of Hierarchical Multilevel Systems” in 1970 1. In addition to describing organisations, the same concept is applicable in the hierarchies of functions and systems. Applying the ideas of decentralised processes and control at hierarchical levels led to the way
Artificial Intelligence – yesterday and today History of Artificial Intelligence (AI) From the early history of Artificial Intelligence a couple of names show up: 1 a Catalan poet and theologian Ramon Llull (1305). He used a paper-built mechanical machine to create new knowledge by combining the existing concepts. 2 a German mathematician and philosopher Gottfried
Big Data and Big Data Analytics Big Data The concept of Big Data originates from the late 90s 1 and it associates in people’s minds with the massive volumes of data that are impossible to process with conventional data processing tools in a reasonable time. However, the volume is not the only feature characterising Big
IoT, CC, and EC – What are these? When explaining the ICT terminology, it is good to keep in mind what Hamilton 1 says while writing about Edge Computing “Like the metaphorical cloud and the Internet of Things, the edge is a buzzword meaning everything and nothing”. IoT is the abbreviation of Internet of Things.
Trends in today’s automation Digitalisation has strongly changed the automation field. This change started already during the mid-60s with the advent of microprocessors, and it has continued with increasing tempo since then. It seems that automation as a separate technology is giving more and more room to information technology that provides the tools and methods
Principal Component Analysis – Example To illustrate the calculations connected to Principal Component Analysis (PCA), let us solve a simple toy problem. Let us assume that we have a small data matrix of 3 variables and 4 experiments, and we want to express all observations using only 2 variables. We first write the data matrix
Principal Component Analysis (PCA) Basic algorithm Raw process data usually includes several variables correlating with each other and increasing the dimensionality of the system or disturbing the regression analysis. Principal Component Analysis, PCA, is a conventional method to decrease this dimensionality without losing the information in the correlated variables. It searches for fewer linear combinations
Capability indices Tolerances and relative precision index In measurement systems, accuracy describes how close the measurement is to its true value. Precision, on the other hand, tells about reproducibility of the measurement: if we repeat the measurement several times, how close the results of repetitions are to each other. In quality control, we can interpret
Cusum charts Cumulative sum (Cusum) chart was developed already during the 1950s for change detection. It mainly applies for detecting the change in the mean value. It is somewhat more complicated to use than the Shewhart charts, but more efficient in detecting small shifts in the mean. This is especially true when the changes are
Control charts Control charts are useful in following the timely development of quality variables. Figure 1 shows the principles of drawing one control chart: the mean control chart. If the process is running satisfactorily, one expects all means of successive samples to lie inside the upper and lower action limits. The probability of a sample
