Smart anomaly detection in sensor systems: A multi-perspective review

Erhan, Laura; Ndubuaku, Maryleen U.; Mauro, Mario Di; Song, Wei; Chen, M.; Fortino, Giancarlo; Bagdasar, Ovidiu; Liotta, Antonio

doi:10.1016/j.inffus.2020.10.001

Cited by 190 publications

(79 citation statements)

References 130 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In some cases, it is hard to find complete information about this kind of characteristic of the data when looking at the statistical and ML-based solutions listed in Section 2 (the surveys in [ 6 ] and [ 14 ] mention potential issues with semi-supervised learning techniques, but do not mention the distributions we may find), so it is not so clear that they are given enough importance (such as in refs. [ 21 , 22 , 52 ], to name a few).…”

Section: Discussionmentioning

confidence: 99%

“…The survey in ref. [ 14 ] also identifies types of problematic datasets and sources of anomalies (the environment, the system, the communication and attacks). It also explains that anomalies have been addressed from several domains: statistical methods, time-series analysis, signal processing, spectral techniques, information theory and machine learning.…”

Section: Related Workmentioning

confidence: 99%

“…The work in ref. [ 14 ] also mentions other approaches in the area of time series analysis, such as auto regressive integrated moving average (ARIMA), but obtaining good models is very complex and problem specific; moreover, when detecting anomalies, they need to assume that the estimation is more accurate than the real value and, therefore, values far from the estimation are considered outliers.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Analyzing Particularities of Sensor Datasets for Supporting Data Understanding and Preparation

Nieto

Aguilera

López-de-Ipiña

2021

Sensors

View full text Add to dashboard Cite

Data scientists spend much time with data cleaning tasks, and this is especially important when dealing with data gathered from sensors, as finding failures is not unusual (there is an abundance of research on anomaly detection in sensor data). This work analyzes several aspects of the data generated by different sensor types to understand particularities in the data, linking them with existing data mining methodologies. Using data from different sources, this work analyzes how the type of sensor used and its measurement units have an important impact in basic statistics such as variance and mean, because of the statistical distributions of the datasets. The work also analyzes the behavior of outliers, how to detect them, and how they affect the equivalence of sensors, as equivalence is used in many solutions for identifying anomalies. Based on the previous results, the article presents guidance on how to deal with data coming from sensors, in order to understand the characteristics of sensor datasets, and proposes a parallelized implementation. Finally, the article shows that the proposed decision-making processes work well with a new type of sensor and that parallelizing with several cores enables calculations to be executed up to four times faster.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Analyzing Particularities of Sensor Datasets for Supporting Data Understanding and Preparation

Nieto

Aguilera

López-de-Ipiña

2021

Sensors

View full text Add to dashboard Cite

show abstract

“…Rūko duomenų koncentratoriaus ar išoriškai surinkti duomenys gali būti naudojami neuroninių tinklų modeliams mokyti. Galimos šios neuroninio tinklo modelio mokymo ir vyk dymo vietos (Erhan et al, 2021):…”

Section: Rūko Kompiuterijos Duomenų Apdorojimo Uždaviniaiunclassified

Fog Computing Technologies for Patient Sensor Networks – Trends, Issues and Future Directions

Kazlauskas

2021

Science - Future of Lithuania

View full text Add to dashboard Cite

Advances in sensors and internet of things promise broad opportunities in many areas and one of them is health care. There are many solutions to manage health care data based on cloud computing. However, high response latency, large volumes of data transferred and security are the main issues of such approach. Fog computing provides immediate response and ways to process large amounts of data using real time analytics which includes machine learning and AI. Fog computing has not yet fully matured and there are still many challenges when managing health care data. It was chosen to investigate the most relevant ehealth fog computing topics by analyzing review articles to explain the fog computing model and present the current trends – fog computing ehealth technology application environments, deployment cases, infrastructure technologies, data processing challenges, problems and future directions. 38 scientific review articles published in the last 5 years were selected for analysis, filtering the most significant works with Web of Science article search tool.

show abstract

“…Its principle is that it looks for anomalies in data or processes so that it can inform maintenance staff in advance of an emerging problem. The fact is that a large number of anomalies in the manufacturing process remains undetected and hidden from conventional detection techniques [ 12 ].…”

Section: Introductionmentioning

confidence: 99%

Smart Anomaly Detection and Prediction for Assembly Process Maintenance in Compliance with Industry 4.0

Tanuška

Spendla

Kebísek

et al. 2021

Sensors

View full text Add to dashboard Cite

One of the big problems of today’s manufacturing companies is the risks of the assembly line unexpected cessation. Although planned and well-performed maintenance will significantly reduce many of these risks, there are still anomalies that cannot be resolved within standard maintenance approaches. In our paper, we aim to solve the problem of accidental carrier bearings damage on an assembly conveyor. Sometimes the bearing of one of the carrier wheels is seized, causing the conveyor, and of course the whole assembly process, to halt. Applying standard approaches in this case does not bring any visible improvement. Therefore, it is necessary to propose and implement a unique approach that incorporates Industrial Internet of Things (IIoT) devices, neural networks, and sound analysis, for the purpose of predicting anomalies. This proposal uses the mentioned approaches in such a way that the gradual integration eliminates the disadvantages of individual approaches while highlighting and preserving the benefits of our solution. As a result, we have created and deployed a smart system that is able to detect and predict arising anomalies and achieve significant reduction in unexpected production cessation.

show abstract

Smart anomaly detection in sensor systems: A multi-perspective review

Cited by 190 publications

References 130 publications

Analyzing Particularities of Sensor Datasets for Supporting Data Understanding and Preparation

Analyzing Particularities of Sensor Datasets for Supporting Data Understanding and Preparation

Fog Computing Technologies for Patient Sensor Networks – Trends, Issues and Future Directions

Smart Anomaly Detection and Prediction for Assembly Process Maintenance in Compliance with Industry 4.0

Contact Info

Product

Resources

About