Performance evaluation of accelerometers ADXL345 and MPU6050 exposed to random vibrational input

The use of vibration monitors is a well-established practice in industrial maintenance, usually vibration sensors are positioned at specific points on the monitored machinery and data are continuously collected to feed a machine operating health control system. Nevertheless, the technology for obtaining the signal, its treatment and analysis is generally expensive, and the financial return is not evident, which justifies the development of low-cost alternatives technologies. In this work was performed an analysis of the responses of two Micro-Electro-Mechanical accelerometers, models ADXL345 and MPU6050, exposed to a low intensity random signal and standard operating frequency. The objective of the analysis was to verify the capacity of these devices to be used as mechanical vibration sensors for rotating machines. For this purpose, offset shift analyzes of the sensors due to the Earth's gravitational field were performed, as well as vibrational spectrum and rectification errors analysis under multiple conditions. The data pointed to a greater uniformity of the MPU6050 response, while several behavioral anomalies were seen in the ADXL345, when these sensors are exposed to the same mechanical signal. The qualitative and quantitative behavior of MPU6050 rectification error was consistent with reported in the literature. It was noted that the methodology used can profile the behavior of sensors, however, it is not sufficient to safely justify the inaccuracies, requiring that the tests be performed on a statistically representative number of sensors from different manufacturers and batches.

Applying ensemble neural networks to analyze industrial maintenance: Influence of Saharan dust transport on gas turbine axial compressor fouling

The planning of industrial maintenance associated with the production of electricity is vital, as it yields a current and future snapshot of an industrial component in order to optimize the human, technical and economic resources of the installation. This study focuses on the degradation due to fouling of a gas turbine in the Canary Islands, and analyzes fouling levels over time based on the operating regime and local meteorological variables. In particular, we study the relationship between degradation and the suspended dust that originates in the Sahara Desert. To this end, we use a computational procedure that relies on a set of artificial neural networks to build an ensemble, using a cross-validated committees approach, to yield the compressor efficiency. The use of trained models makes it possible to know in advance how the local fouling of an industrial rotating component will evolve, which is useful for maintenance planning and for calculating the relative importance of the variables that make up the system

A Modular Federated Learning Architecture for Integration of AI-enhanced Assistance in Industrial Maintenance - A novel architecture for enhancing industrial maintenance management systems in the automotive and semiconductor industry.

Artificial Intelligence (AI) plays an increasingly important role for the implementation and failure-free operation of Cyber-Physical Production Systems (CPPS). Recent market studies show that investment in AI-enhanced maintenance is increasing as one of the most important use cases of Industry 4.0. AI systems enable the improvement of various Key Performance Indicators (KPI), ultimately leading to a reduction in costs and optimizing plant management in smart factories. At the same time, manufacturing enterprises in diverse sectors have very high expectations from any kind of AI solution comparing to conventional solutions. Today manufacturing enterprises use only a quarter of their data and therefore leave an enormous, untapped potential. The use of Text Mining (TM) realizes the untapped value of existing unstructured or semi-structured textual data. This paper presents a transferable and scalable architecture for a cognitive maintenance system of a human-centered assistance system that enables holistic sensing of the environment by using physical and virtual sensors. By focusing on generalizability, scalability, adaptability, reliability, and user acceptance, a novel architecture for cognitive maintenance system is proposed. The so called ARCHIE, Architecture for a Cognitive Maintenance System, addresses common challenges in the application of AI systems in the industrial environment. Human-centered cognitive systems aim to automate manufacturing processes and assist workers in their cognitive tasks. This can be achieved by using the untapped potential of combining unstructured and structured data in order to extract hidden knowledge. ARCHIE aims at realizing an AI-enhanced approach for a human-centered assistance system. ARCHIE incorporates physical and virtual sensors that capture machine states, parameters, human knowledge, and skills to optimize relevant KPIs. This includes a reduction in documentation time, Mean Time Between Failures (MTBF) and Mean Failure Detection Time (MFDT), as well as an increase in uptime, leading ultimately to an improved Overall Equipment Efficiency (OEE). These improvements are enabled by the combined use of AI in the form of TM, Federated Learning and Knowledge Graphs. In the presented use-case from the automotive industry, a reduction in MFDT below 60min by 97.3% and an increase in OEE by 5.3% was achieved. In the Semiconductor industry, the partial application of ARCHIE allows the querying of competence distributions based on a given maintenance task, enabling automated allocation of maintenance technicians and trend analyses. Generalizability, scalability, adaptability, reliability, and user acceptance were also evaluated in the use cases presented.

On the optimization of maintenance storage cost in industry a fuzzy logic application

PurposeThe purpose of this paper is to construct a fuzzy logic model that acts as a decision support system to minimize inventory-related costs in the field of industrial maintenance. Achieving a balance between the unavailability and over-storage of spare parts is a problem with potentially significant consequences. That significance increases proportionally with the ever-increasing challenge of reducing overall cost. Either scenario can result in substantial financial losses because of the interruption of production or the costs of tied-up capital, also called the “solidification of capital.” Moreover, there is that additional problem of the expiry of parts on the shelf.Design/methodology/approachThe proposed approach relies on inputs from experts with consideration for incompleteness and inaccuracy. Two levels of decision are considered simultaneously. The first is whether a part should be stored or ordered when needed. The second involves comparing suppliers with their batch-size offers based on user-determined criteria. A mathematical model is developed in parallel for validation.FindingsThe results indicate that the fuzzy logic approach is accurate and satisfactory for this application and that it is advantageous because of its limited sensitivity to the inaccuracy and/or incompleteness of data. In addition, the approach is practical because it requires minimal user effort.Originality/valueTo the best of the authors’ knowledge, the exploitation of fuzzy-logic altogether with limited sensitivity experts' inputs were never combined for the solution of this particular problem; however, this approach's positive impact is expected to be highly significant in solving a chronic problem in industry.

Developing a Data-Driven Unsupervised Pattern Recognition Approach for Sensor Signal Anomaly Detection

Coming up with a system for early detection of machine damages and failures is one of the important challenges in the industrial maintenance procedure to avoid additional costs and downtimes. To approach this goal, this paper uses the signal gathered by a sensing system which employed a spintropic sensor to measure the magnetic field around the machine which somehow shows the machine's behaviour. Using this signal and focusing on analysing and processing the signal, this paper develops a data-driven method to recognize signal patterns and subsequently detects anomalies. A challenging task that we succeeded to overcome in this paper is recognizing relevant signal patterns without having any prior knowledge. An algorithm designed for this task is therefore completely unsupervised which makes it consistent and suitable to apply it for the signals gathered for other types of machines. Using both frequency and time domain information, the proposed algorithm, which utilizes signal processing and machine learning techniques, is able to efficiently identify relevant signal patterns. Clustering results on the real data gathered by the aforementioned sensor have shown the high accuracy of 99.38% in recognizing patterns. Furthermore, an anomaly score measure is used and according to its distribution, anomalies are detected appropriately. <br>

Gone in 2s: a deep dive into perfection analysing the collaborative maintenance pitstop of Formula 1

PurposeThe paper aims to convert into useable guidelines, the knowledge related to human factors and tasks' organisation, which are embedded in one of the most exciting maintenance actions that are carried out, the pitstop in Formula 1 races.Design/methodology/approachThe paper opted for a fault tree analysis (FTA) to de-construct all the sub-tasks and their possible deviations from desirable situations and to evaluate the most relevant information needed for carrying out the pitstop operation. Besides, the SHELL model was applied in a second stage to evaluate the interaction between human being and human interfaces with other components of the system. Once this set of information was crystallised, the research translated it into useable guidelines for organising industrial maintenance actions using the same approach and possible reaching the same results.FindingsThe results of this study is a structured set of guidelines that encompasses the most paramount aspects that should be considered for setting correct maintenance actions. They represent a “guide” for including the different angles that are included during these operations.Research limitations/implicationsThe guidelines are potentially applicable to every maintenance operation. The guidelines should be tested on different working domains to check their applicability besides the racing world.Practical implicationsThis study is a reverse engineering work for creating a scheme to include into maintenance operations aspects such as crew athlete-like fitness, training, technology, organisational issues, safety, ergonomics and psychology.Originality/valueThe value of the paper is deconstructing the results of one of the most successful and prepared maintenance action. The paper takes a different approach in proposing how to structure and create maintenance solutions. The difference in approaches between the maintenance during the pitstop of Formula 1 car and industrial applications enhances the gap that needs still to be filled for further improving maintenance actions out of the racing world.