Smart data driven maintenance: Improving damage detection and assessment on aerospace structures

Archetti, Francesco; Arosio, G; Candelieri, Antonio; Giordani, Ilaria; Sormani, Raul

doi:10.1109/metroaerospace.2014.6865902

Cited by 6 publications

(5 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The prediction models are generated as code and applied to the datastream of the sensor readings to obtain real-time prediction of RUL [10]. The prediction results are communicated with the CMMS via e-mail or customized dashboards.…”

Section: Ai Modulementioning

confidence: 99%

“…The model is trained to predict the value of the engineered binary parameter, thus the problem is a two-class classification problem. The output of the model is the probability that RUL[10 cycles, hereinafter denoted as P r (RUL [10). After iterations with logistic regression, neural network, decision forest, and decision jungle, it is found that the two-class neural network gives the best prediction model for the turbofan jet engine run-to-failure dataset.…”

Section: Ai Modulementioning

confidence: 99%

“…To demonstrate the results, P r (RUL [10) for the entire lifetime of one engine is plotted in Fig.10. In addition, Fig.…”

Section: Ai Modulementioning

confidence: 99%

“…Currently, it seems that DeepMind is able to learn more like a human; the program can learn tasks sequentially using skills it acquires on the way [9]. For maintenance, AI is considered to be a data-driven method, where an artificial neural network (ANN) is applied to estimate the remaining useful life [10]. In the context of process industry, a literature study has been conducted on data-driven methods [6].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Deep digital maintenance

2017

View full text Add to dashboard Cite

With the emergence of Industry 4.0, maintenance is considered to be a specific area of action that is needed to successfully sustain a competitive advantage. For instance, predictive maintenance will be central for asset utilization, service, and after-sales in realizing Industry 4.0. Moreover, artificial intelligence (AI) is also central for Industry 4.0, and offers data-driven methods. The aim of this article is to develop a new maintenance model called deep digital maintenance (DDM). With the support of theoretical foundations in cyber-physical systems (CPS) and maintenance, a concept for DDM is proposed. In this paper, the planning module of DDM is investigated in more detail with realistic industrial data from earlier case studies. It is expected that this planning module will enable integrated planning (IPL) where maintenance and production planning can be more integrated. The result of the testing shows that both the remaining useful life (RUL) and the expected profit loss indicator (PLI) of ignoring the failure can be calculated for the planning module. The article concludes that further research is needed in testing the accuracy of RUL, classifying PLI for different failure modes, and testing of other DDM modules with industrial case studies.

show abstract

Section: Ai Modulementioning

confidence: 99%

Section: Ai Modulementioning

confidence: 99%

“…To demonstrate the results, P r (RUL [10) for the entire lifetime of one engine is plotted in Fig.10. In addition, Fig.…”

Section: Ai Modulementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Deep digital maintenance

2017

View full text Add to dashboard Cite

show abstract

“…As an example of artificial intelligence could be application of Artificial Neural Network (ANN) that has been successfully applied for aerospace structures [19] and Wind Energy Conversion System [20].…”

Section: Data-driven Predictive Maintenancementioning

confidence: 99%

Data-driven Predictive Maintenance for Green Manufacturing

Rødseth¹,

Schjølberg²

2016

Proceedings of the 6th International Workshop of Advanced Manufacturing and Automation

View full text Add to dashboard Cite

Abstract-With the current situation of high demand of sustainable manufacturing, different stakeholders have clear expectations for more environmental manufacturing and at the same time minimizing the operational costs. The role of maintenance plays a key role in the path towards sustainable manufacturing. For achieving green manufacturing, more datadriven predictive maintenance strategies is needed and is expected to reduce energy consumption, maintenance resources in terms of spare parts, and reduction of consumables in terms of example lubrication. The overall bottom-line for the predictive maintenance strategy is increased availability, reduction of maintenance hours in terms of reactive maintenance activities, and increased profit for the manufacturing business. For a predictive maintenance strategy, it is crucial to develop Key Performance Indicators (KPIs) for the maintenance management. Today, common KPIs such as availability and different indicators for maintenance cost has been developed. When aiming for more green manufacturing, a more integrated application of maintenance KPIs are needed. Today, the KPI Profit Loss Indicator (PLI) has been developed and demonstrated in the saw mill industry and is regarded to support a more integrated approach in terms of Integrated Planning (IPL). The aim of this article is develop a structured approach for data-driven predictive maintenance aligned with the concept of PLI. Through a case study, the approach is partly demonstrated for the manufacturing industry. The results in this demonstration shows that the data-driven maintenance strategy will have a positive impact of the PLI value and provide a sustainable manufacturing in long-term.

show abstract