Creation of Publicly Available Data Sets for Prognostics and Diagnostics Addressing Data Scenarios Relevant to Industrial Applications

Abstract

For a successful realization of prognostics and health management (PHM), the availability of sufficient run-to-failure data sets is a crucial factor. The sheer number of given data points holds less importance than the full coverage of the potential state space. However, full coverage is a major challenge in most industrial applications. Among other things, high investment and operating costs as well as the long service life of many technical systems make it difficult to acquire complete run-to-failure data sets. Consequently, in industrial applications data sets with specific deficiencies are frequently encountered. The development of appropriate methods to address such data scenarios is a fundamental research issue. Therefore, the purpose of this paper is to provide facilitation for this research. Accordingly, the paper starts by specifying the value and availability of data in PHM. Subsequently, criteria for characterizing data sets are defined independent of the actual PHM application. The criteria are used to identify typical data scenarios with specific deficiencies that possess significant relevance for industrial applications. Thereafter, the most comprehensive overview of data sets suitable for PHM and currently publicly accessible is provided. Thereby, not all previously identified data scenarios with their specific deficiencies are addressed by at least one data set. A program is established for the aforementioned facilitation of further research. One objective of the program is to create data sets reflecting these data scenarios using a test bench. First, possible applications and their degradation processes to be studied on the test bench are briefly characterized. Thereby, the final decision to select filtration as a test bench application is argued. Subsequently, the test bench created is introduced, including a description of the functional concept, pneumatic layout and components involved, as well as the filter media and test dusts employed. Typical run-to-failure trajectories are illustrated. Thereafter, the data set published under the name Preventive to Predictive Maintenance is presented. Additionally, a schedule for future releases of data sets on further industry-relevant data scenarios is sketched.

Aplicação de técnicas de mineração em dados sintéticos para manutenção preditiva: um estudo de caso

“…De acordo com [Mauthe et al 2021], a baixa disponibilidade de dados reais de falhas é um dos grandes desafios de pesquisa na área de PHM. Os autores analisam datasets públicos de manutenc ¸ão preditiva, concluindo que a vasta maioria advêm de simulac ¸ões e ensaios de laboratório e alertam que a maioria não cobre de forma significativa todo o espac ¸o de estados relevante para a previsão de falhas.…”

“…A construc ¸ão de modelos preditivos PHM tem sido intensamente estudada do ponto de vista da construc ¸ão dos modelos usando técnicas de aprendizagem de máquina [Zhang et al 2019, Susto et al 2015, Mathew et al 2017, Carchiolo et al 2019, Mahmoodzadeh et al 2020. Contudo, para que retratem fidedignamente o processo de degradac ¸ão das máquinas, dependem de dados reais de operac ¸ão e falhas destes ativos, o que representa atualmente um grande desafio na área [Mauthe et al 2021]. Tal dificuldade é devida principalmente a questões estratégicas das empresas e fabricantes de equipamentos, que não desejam que tais dados se tornem públicos.…”

“…Neste cenário, a gerac ¸ão de dados sintéticos se apresenta como uma soluc ¸ão técnica e economicamente viável para construc ¸ão de modelos preditivos para PHM. Porém, para construc ¸ão de um modelo de PHM eficaz para os cenários industriais relevantes deve-se considerar uma série de requisitos do conjunto de dados, entre eles a cobertura do espac ¸o de estados possível das variáveis [Mauthe et al 2021]. Logo, a simples gerac ¸ão de dados por simuladores pode não ser suficientemente verossímil para a construc ¸ão de modelos preditivos úteis em cenários práticos da indústria.…”

Remaining Useful Life Prediction using Gaussian Process Regression Model

“…Finally, the number of perceptions in the output layer is equivalent to the number of output parameters. To illustrate the performance of the RUL predictions, analysis is performed on the filter clogging dataset 1 (Mauthe et al, 2021). Almost all industries depend on some sort of filtration process and filter clogging tends to be one of the main reasons for filter degradation and removal (Sutherland & Chase, 2011).…”

“…To illustrate the benefits of the GPR model, the RUL of dust filters are examined (Mauthe, Hagmeyer, & Zeiler, 2021). These types of filters are utilized in a variety of industries such as automotive, aviation, and facility management.…”