1,721,368 research outputs found
A Neural Network Approach to Find The Cumulative Failure Distribution: Modeling and Experimental Evidence
No full textThe failure prediction of components plays an increasingly important role in manufacturing. In this context, new models are proposed to better face this problem, and, among them, artificial neural networks are emerging as effective. A first approach to these networks can be complex, but in this paper, we will show that even simple networks can approximate the cumulative failure distribution well. The neural network approach results are often better than those based on the most useful probability distribution in reliability, the Weibull. In this paper, the performances of multilayer feedforward basic networks with different network configurations are tested, changing different parameters (e.g., the number of nodes, the learning rate, and the momentum). We used a set of different failure data of components taken from the real world, and we analyzed the accuracy of the approximation of the different neural networks compared with the least squares method based on the Weibull distribution. The results show that the networks can satisfactorily approximate the cumulative failure distribution, very often better than the least squares method, particularly in cases with a small number of available failure times
Forecasting methods for lumpy demand of aircraft spare parts
No full textAn accurate demand forecasting is a critical issue of the industrial plants management.
This study analyses the behaviour of forecasting techniques when dealing with lumpy demand, measured by the square coefficient of variation (CV) and the average inter-demand interval (ADI).
In particular different forecasting techniques are considered: actual historical data from the Italian national flag airline are used for their performance analysis and comparison.
This study demonstrate that the item lumpiness is a dominant parameter. The results attest that demand forecasting for lumpy items is a very complex problem and results obtained by existing approaches are not very accurate. Anyway, the Seasonal Regression Model (SRM), the Exponentially Weighted Moving Average (EWMA(i)) and Winters model reveal the best approaches for the prediction of spare pat-is demand of airline fleet
Automatic assessment of the ergonomic risk for manual manufacturing and assembly activities through optical motion capture technology
No full textSafeguard the operator health is nowadays a hot topic for most of the companies whose production process relies on manual manufacturing and assembly activities. European legislations, national regulations and international standards force the companies to assess the risk of musculoskeletal disorders of operators while they are performing manual tasks. Furthermore, international corporates typically require their partners to adopt and implement particular indices and procedures to assess the ergonomic risks specific of their industrial sector. The expertise and time required by the ergonomic assessment activity compels the companies to huge financial, human and technological investments. An original Motion Analysis System (MAS) is developed to facilitate the evaluation of most of the ergonomic indices traditionally adopted by manufacturing firms. The MAS exploits a network of marker-less depth cameras to track and record the operator movements and postures during the performed tasks. The big volume of data provided by this motion capture technology is employed by the MAS to automatically and quantitatively assesses the risk of musculoskeletal disorders over the entire task duration and for each body part. The developed hardware/software architecture is tested and validated with a real industrial case study of a car manufacturer which adopts the European Assembly Worksheet (EAWS) to assess the ergonomic risk of its assembly line operators. The results suggest how the MAS is a powerful architecture compared to other motion capture solutions. Indeed, this technology accurately assesses the operator movements and his joint absolute position in the assembly station 3D layout. Finally, the MAS automatically and quantitatively fill out the different EAWS sections, traditionally evaluated through time- and resource-consuming activities
Prognostic Health Management of Production Systems. New Proposed Approach and Experimental Evidences
Full text linkPrognostic Health Management (PHM) is a maintenance policy aimed at predicting the occurrence of a failure in components in order to minimize unexpected downtimes of complex systems and maximize their availability. Recent developments in condition monitoring (CM) techniques and Artificial Intelligence (AI) tools enabled the collection of a huge amount of data in real-time and its transformation into meaningful information that will support the maintenance decision-making process. The emerging Cyber-Physical Systems (CPS) technologies connect distributed physical systems with their virtual representations in the cyber computational world. The PHM assumes a key role in the implementation of CPS in manufacturing contexts, since it allows to keep CPS and its machines in proper conditions. On the other hand, CPS-based PHM provide an efficient solution to maximize availability of machines and production systems.
In this paper, evolving and unsupervised approaches for the implementation of PHM at a component level are described, which are able to process streaming data in real-time and with almost-zero prior knowledge about the monitored component. A case study from a real industrial context is presented. Different unsupervised and online anomaly detection methods are combined with evolving clustering models in order to detect anomalous behaviors in streaming vibration data and integrate the so-generated knowledge into supervised and adaptive models; then, the degradation model for each identified fault is built and the resulting RUL prediction model integrated into the online analysis. Supervised methods are applied to the same dataset, in batch mode, to validate the proposed procedure
New Kanban model for tow-train feeding system design
No full textPurpose – This paper aims to introduce, apply and validate, through a realistic case study, an analytical cost model to support the design of the
tow-train feeding system for mixed-model assembly lines managed according to the just-in-time concept. The fleet size and inventory level,
minimizing the total annual cost, are the key model goals, while the tow-train shipping capacity and the service level are the decisional variables
to set.
Design/methodology/approach – The model computes the material handling, inventory and stockout rising costs of the tow-train feeding system
and looks for their minimization. It further computes the expected lead time between consecutive round-trips and the Kanban card number,
distinguishing among parts and assembly lines, overcoming the simplifying hypothesis assuming a constant lead time for all parts. The model is
validated against a dedicated case study stressing its strengths in terms of cost and inventory-level reduction.
Findings – The proposed approach is found to be effective if compared to the standard literature in the field of Kanban system design. The 10.76
per cent cost saving is experienced for the considered case study, and the inventory level is closer to the field-experienced profile.
Practical implications – The model adopts a practical perspective, making it easy and applicable to common operative industries.
Originality/value – The literature neglects to consider the differences in the part consumption when estimating the lead time between tow-train
round-trips. The proposed model overcomes such limitations and strengthens the model applicability and performances
Operations, Logistics and Supply Chain Management: Definitions and Objectives
No full textOperations, logistics and supply chains are catalysts in any modern economy and therefore essential contributors to economic prosperity and societal welfare. This chapter briefly sketches the origins of the field and presents a case study on the importance of a balanced logistical organization from the 17th century, after which formal definitions and objectives are introduced. In addition, we discuss relations with other management areas as well as with other science domains such as law or social and political sciences. Topics and concepts in this chapter are discussed at an elementary level, aiming to provide an introduction to the topical field of operations, logistics, and supply chain management
Artificial neural network optimisation for monthly average daily global solar radiation prediction
No full textThe availability of reliable climatologic data is essential for multiple purposes in a wide set of anthropic activities and operative sectors. Frequently direct measures present spatial and temporal lacks so that predictive approaches become of interest. This paper focuses on the prediction of the Monthly Average Daily Global Solar Radiation (MADGSR) over Italy using Artificial Neural Networks (ANNs). Data from 45 locations compose the multi-location ANN training and testing sets. For each location, 13 input parameters are considered, including the geographical coordinates and the monthly values for the most frequently adopted climatologic parameters. A subset of 17 locations is used for ANN training, while the testing step is against data from the remaining 28 locations. Furthermore, the Automatic Relevance Determination method (ARD) is used to point out the most relevant input for the accurate MADGSR prediction. The ANN best configuration includes 7 parameters, only, i.e. Top of Atmosphere (TOA) radiation, day length, number of rainy days and average rainfall, latitude and altitude. The correlation performances, expressed through statistical indicators as the Mean Absolute Percentage Error (MAPE), range between 1.67% and 4.25%, depending on the number and type of the chosen input, representing a good solution compared to the current standards
Optimal design of AS/RS storage systems with three-class-based assignment strategy under single and dual command operations
No full textThis paper presents an extension of the analytical
models already proposed by the literature to compute the expected
travel time of automated storage and retrieval systems
(AS/RS) in three-class-based storage (3-CBS) rectangular-intime
(RIT) storage systems. The authors determined the analytical
closed form of the mean travel time for both the singlecommand
(SC) and the dual-command (DC) cycles varying
the warehouse shape factor and the ABC turnover curve. The
performances obtained by the adoption of the proposed analytical
travel time model under different configurations of the
warehousing system, i.e., shape, dimension of the classes, and
ABC curve, are evaluated and compared. Finally, the optimal
boundary limits for the 3-CBS AS/RS, considering both the
SC and the DC cycles, are fixed presenting the percentage
saving of such configurations toward the common random
assignment policy
Components monitoring and intelligent diagnosis tools for Prognostic Health Management approach
No full textThe main goal of maintenance of complex systems is to minimize downtimes to make the system as much available as possible. Condition-Based Maintenance (CBM) is one of the most effective policies used by Companies nowadays, based on the monitoring of different parameters of machines that reflect its health status. CBM can be implemented by using the Prognostic Health Management approach, made up of four main steps: data collection, signal processing, diagnostic, and prognostic. It is a proactive process that requires the development of predictive models that can trigger the alarm for corresponding maintenance. The huge amount of data that need to be collected has suggested the use of models coming from statistic theory and data mining, in order to discover regular pattern in large data sets and generate knowledge that will be useful in the maintenance decision-making process. In this paper, different intelligent methods for diagnostic purpose, such as Decision trees, K-NN algorithm, Artificial Neural Networks and support Vector Machine, are used to classify the health condition of a rotating component. Collected signals are processed in the time-domain and in the time-frequency-domain in order to extract relevant features to give as input data for the intelligent methods. Such methods are finally compared by evaluating the related accuracy value for both training and testing. The main result of this work is that the time-frequency analysis improves accuracy in classifying the health condition of machines and that new intelligent models can perform in an effective way even in the time-domai
Sustainability in the tobacco supply chain through low-carbon flue curing processes
No full textThis study addresses the goal of greening the tobacco supply chain, presenting a joint technical, economic and environmental feasibility study to reduce the impact of the high-energy intensive tobacco flue-curing phase, aimed at drying the tobacco leaves. Starting from a review of the process, presenting its standard 1-week cycle, the curing barn features and the required physical conditions to obtain top quality flue-cured tobacco, this study investigates the impact of non-fossil fuels feeding the heat generators, e.g. pellets, woodchips. The feasibility study input data are from Italian producers, located in the Umbria region selling tobacco to an international leading company that joined the present research. Results highlight that the switch to non-fossil fuels for tobacco curing may lead to annual cost savings up to 13% and to global environmental saving up to 95% without any product quality decrease
- …
