Communications in computer and information science, 2015
Today sensors are widely used in many monitoring applications. Due to some random environmental e... more Today sensors are widely used in many monitoring applications. Due to some random environmental effects and/or sensing failures, the collected sensor data is typically noisy. Thus, it is critical to cleanse the data before using it for answering queries or for data analysis. Popular data cleansing approaches, such as classification, prediction and moving average, are not suited for embedded sensor devices, due to their limit storage and processing capabilities. In this paper, we propose a sensor data cleansing approach using the relational-based technologies, including constraints, triggers and granularity-based data aggregation. The proposed approach is simple but effective to cleanse different types of dirty data, including delayed data, incomplete data, incorrect data, duplicate data and missing data. We evaluate the proposed strategy to verify its efficiency and effectiveness.
With the prevalence of cloud computing and Internet of Things (IoT), smart meters have become one... more With the prevalence of cloud computing and Internet of Things (IoT), smart meters have become one of the main components of smart city strategies. Smart meters generate large amounts of fine-grained data that is used to provide useful information to consumers and utility companies for decision making. Now-a-days, smart meter analytics systems consist of analytical algorithms that process massive amounts of data. These analytics algorithms require ample amounts of realistic data for testing and verification purposes. However, it is usually difficult to obtain adequate amounts of realistic data, mainly due to privacy issues. This paper proposes a smart meter data generator that can generate realistic energy consumption data by making use of a small real-world data set as seed. The generator generates data using a prediction-based method that depends on historical energy consumption patterns along with Gaussian white noise. In this paper, we comprehensively evaluate the efficiency and effectiveness of the proposed method based on a real-world energy data set.
• Users may download and print one copy of any publication from the public portal for the purpose... more • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal Download policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Artificial intelligence allows small and medium-sized enterprises (SMEs) in the manufacturing sec... more Artificial intelligence allows small and medium-sized enterprises (SMEs) in the manufacturing sector to improve performance, reduce downtime and increase productivity. SMEs in Denmark are still struggling to implement artificial intelligence based strategies since they face a range of challenges, such as business applications, data availability, organizational culture towards the acceptance of new technologies, investment in new technologies, skills gap, development process and effective strategy. In the beginning, the paper describes the challenges faced by SMEs in adopting artificial intelligence. Then, the paper suggests solutions to overcome these challenges and discusses the importance of artificial intelligence as well as the opportunities it
Time series data grouping is essential for empirical data analysis and data summarization. Genera... more Time series data grouping is essential for empirical data analysis and data summarization. Generally, grouping of time series data is based on similarity measure (distance function), thus time series in the same group are similar. The choice of similarity measure is very important during data grouping. In this paper we investigate the issue of smart meter data grouping according to daily consumption pattern using either Euclidian or correlation-based similarity. We find that the correlation-based measure is superior for data grouping with respect to consumption pattern.
• Users may download and print one copy of any publication from the public portal for the purpose... more • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal Download policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Abstract Analyzing sensor data from a production environment is quite challenging because of the ... more Abstract Analyzing sensor data from a production environment is quite challenging because of the high-dimensional nature of the data. In addition, the generated data is in the form of time-series, where the sequence of registrations may be of utmost significance. One of the main goals of the paper is to determine if the given time-series of feature combinations is normal or rare. This goal could successfully be achieved by combining multiple machine learning models. In this paper, a sliding window based ensemble method is proposed to detect outliers in a streaming fashion. The proposed method uses a combination of clustering algorithms to construct subgroups (clusters) representing different data structures. These structures are later used in a one-class classification algorithm to identfy the outliers. Thus, if a pattern does not belong to any of the common structures or clusters, it is an outlier. Further, based on the rare pattern classification, machine failures could be predicted in advance.
Artificial intelligence allows small and medium-sized enterprises (SMEs) in the manufacturing sec... more Artificial intelligence allows small and medium-sized enterprises (SMEs) in the manufacturing sector to improve performance, reduce downtime and increase productivity. SMEs in Denmark are still struggling to implement artificial intelligence based strategies since they face a range of challenges, such as business applications, data availability, organizational culture towards the acceptance of new technologies, investment in new technologies, skills gap, development process and effective strategy. In the beginning, the paper describes the challenges faced by SMEs in adopting artificial intelligence. Then, the paper suggests solutions to overcome these challenges and discusses the importance of artificial intelligence as well as the opportunities it
Communications in computer and information science, 2020
The technological revolution, known as industry 4.0, aims to improve efficiency/productivity and ... more The technological revolution, known as industry 4.0, aims to improve efficiency/productivity and reduce production costs. In the Industry 4.0 based smart manufacturing environment, machine learning techniques are deployed to identify patterns in live data by creating models using historical data. These models will then predict previously undetectable incidents. This paper initially performs a descriptive statistics and visualization, subsequently issues like classification of data with imbalanced class distribution are addressed. Then several binary classification-based machine learning models are built and trained for predicting production line disruptions, although only logistic regression and artificial neural networks are discussed in detail. Finally, it evaluates the effectiveness of the machine learning models as well as the overall utilization of the manufacturing operation in terms of availability, performance and quality.
Communications in computer and information science, 2015
Today sensors are widely used in many monitoring applications. Due to some random environmental e... more Today sensors are widely used in many monitoring applications. Due to some random environmental effects and/or sensing failures, the collected sensor data is typically noisy. Thus, it is critical to cleanse the data before using it for answering queries or for data analysis. Popular data cleansing approaches, such as classification, prediction and moving average, are not suited for embedded sensor devices, due to their limit storage and processing capabilities. In this paper, we propose a sensor data cleansing approach using the relational-based technologies, including constraints, triggers and granularity-based data aggregation. The proposed approach is simple but effective to cleanse different types of dirty data, including delayed data, incomplete data, incorrect data, duplicate data and missing data. We evaluate the proposed strategy to verify its efficiency and effectiveness.
With the prevalence of cloud computing and Internet of Things (IoT), smart meters have become one... more With the prevalence of cloud computing and Internet of Things (IoT), smart meters have become one of the main components of smart city strategies. Smart meters generate large amounts of fine-grained data that is used to provide useful information to consumers and utility companies for decision making. Now-a-days, smart meter analytics systems consist of analytical algorithms that process massive amounts of data. These analytics algorithms require ample amounts of realistic data for testing and verification purposes. However, it is usually difficult to obtain adequate amounts of realistic data, mainly due to privacy issues. This paper proposes a smart meter data generator that can generate realistic energy consumption data by making use of a small real-world data set as seed. The generator generates data using a prediction-based method that depends on historical energy consumption patterns along with Gaussian white noise. In this paper, we comprehensively evaluate the efficiency and effectiveness of the proposed method based on a real-world energy data set.
• Users may download and print one copy of any publication from the public portal for the purpose... more • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal Download policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Artificial intelligence allows small and medium-sized enterprises (SMEs) in the manufacturing sec... more Artificial intelligence allows small and medium-sized enterprises (SMEs) in the manufacturing sector to improve performance, reduce downtime and increase productivity. SMEs in Denmark are still struggling to implement artificial intelligence based strategies since they face a range of challenges, such as business applications, data availability, organizational culture towards the acceptance of new technologies, investment in new technologies, skills gap, development process and effective strategy. In the beginning, the paper describes the challenges faced by SMEs in adopting artificial intelligence. Then, the paper suggests solutions to overcome these challenges and discusses the importance of artificial intelligence as well as the opportunities it
Time series data grouping is essential for empirical data analysis and data summarization. Genera... more Time series data grouping is essential for empirical data analysis and data summarization. Generally, grouping of time series data is based on similarity measure (distance function), thus time series in the same group are similar. The choice of similarity measure is very important during data grouping. In this paper we investigate the issue of smart meter data grouping according to daily consumption pattern using either Euclidian or correlation-based similarity. We find that the correlation-based measure is superior for data grouping with respect to consumption pattern.
• Users may download and print one copy of any publication from the public portal for the purpose... more • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal Download policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Abstract Analyzing sensor data from a production environment is quite challenging because of the ... more Abstract Analyzing sensor data from a production environment is quite challenging because of the high-dimensional nature of the data. In addition, the generated data is in the form of time-series, where the sequence of registrations may be of utmost significance. One of the main goals of the paper is to determine if the given time-series of feature combinations is normal or rare. This goal could successfully be achieved by combining multiple machine learning models. In this paper, a sliding window based ensemble method is proposed to detect outliers in a streaming fashion. The proposed method uses a combination of clustering algorithms to construct subgroups (clusters) representing different data structures. These structures are later used in a one-class classification algorithm to identfy the outliers. Thus, if a pattern does not belong to any of the common structures or clusters, it is an outlier. Further, based on the rare pattern classification, machine failures could be predicted in advance.
Artificial intelligence allows small and medium-sized enterprises (SMEs) in the manufacturing sec... more Artificial intelligence allows small and medium-sized enterprises (SMEs) in the manufacturing sector to improve performance, reduce downtime and increase productivity. SMEs in Denmark are still struggling to implement artificial intelligence based strategies since they face a range of challenges, such as business applications, data availability, organizational culture towards the acceptance of new technologies, investment in new technologies, skills gap, development process and effective strategy. In the beginning, the paper describes the challenges faced by SMEs in adopting artificial intelligence. Then, the paper suggests solutions to overcome these challenges and discusses the importance of artificial intelligence as well as the opportunities it
Communications in computer and information science, 2020
The technological revolution, known as industry 4.0, aims to improve efficiency/productivity and ... more The technological revolution, known as industry 4.0, aims to improve efficiency/productivity and reduce production costs. In the Industry 4.0 based smart manufacturing environment, machine learning techniques are deployed to identify patterns in live data by creating models using historical data. These models will then predict previously undetectable incidents. This paper initially performs a descriptive statistics and visualization, subsequently issues like classification of data with imbalanced class distribution are addressed. Then several binary classification-based machine learning models are built and trained for predicting production line disruptions, although only logistic regression and artificial neural networks are discussed in detail. Finally, it evaluates the effectiveness of the machine learning models as well as the overall utilization of the manufacturing operation in terms of availability, performance and quality.
Uploads
Papers by Finn Nordbjerg