Fault identification is one of the most significant bottlenecks faced by electricity transmission... more Fault identification is one of the most significant bottlenecks faced by electricity transmission and distribution utilities in developing countries to deliver efficient services to the customers and ensure proper asset audit and management for network optimization and load forecasting. This is due to data scarcity, asset inaccessibility and insecurity, ground-surveys complexity, untimeliness, and general human cost. In view of this, we exploited the use of oblique UAV imagery with a high spatial resolution and a fine-tuned and deep Convolutional Neural Networks (CNNs) to monitor four major Electric power transmission network (EPTN) components. This study explored the capability of the Single Shot Multibox Detector (SSD), a one-stage object detection model on the electric transmission power line imagery to localize, detect and classify faults. The fault considered in this study include the broken insulator plate, missing insulator plate, missing knob, and rusty clamp. Our adapted ne...
Understanding the biases in Deep Neural Networks (DNN) based algorithms is gaining paramount impo... more Understanding the biases in Deep Neural Networks (DNN) based algorithms is gaining paramount importance due to its increased applications on many real-world problems. A known problem of DNN penalizing the underrepresented population could undermine the efficacy of development projects dependent on data produced using DNN-based models. In spite of this, the problems of biases in DNN for Land Use and Land Cover Classification (LULCC) have not been a subject of many studies. In this study, we explore ways to quantify biases in DNN for land use with an example of identifying school buildings in Colombia from satellite imagery. We implement a DNN-based model by fine-tuning an existing, pre-trained model for school building identification. The model achieved overall 84% accuracy. Then, we used socioeconomic covariates to analyze possible biases in the learned representation. The retrained deep neural network was used to extract visual features (embeddings) from satellite image tiles. The ...
Computer vision for large scale building detection can be very challenging in many environments a... more Computer vision for large scale building detection can be very challenging in many environments and settings even with recent advances in deep learning technologies. Even more challenging is modeling to detect the presence of specific buildings (in this case schools) in satellite imagery at a global scale. However, despite the variation in school building structures from rural to urban areas and from country to country, many school buildings have identifiable overhead signatures that make them possible to be detected from high-resolution imagery with modern deep learning techniques. Our hypothesis is that a Deep Convolutional Neural Network (CNN) could be trained for successful mapping of school locations at a regional or global scale from high-resolution satellite imagery. One of the key objectives of this work is to explore the possibility of having a scalable model that can be used to map schools across the globe. In this work, we developed AI-assisted rapid school location mappi...
The study of the dynamic relationship between topological structure of a transit network and the ... more The study of the dynamic relationship between topological structure of a transit network and the mobility patterns of transit vehicles on this network is critical towardsdevising smart and time-aware solutions to transit management and recommendation systems. This paper proposes a time-varying graph (TVG) to model thisrelationship. The effectiveness of this proposed model has been explored by implementing the model in Neo4j graph database using transit feeds generated by bus transit network of the City of Moncton, New Brunswick, Canada. Dynamics in this relationshipalsohave been detected using network metrics such as temporal shortest paths, degree, betweenness and PageRank centralities as well as temporal network diameter and density. Keywords: Transit Networks,Mobility Pattern,Time-Varying Graph model, Graph Databaseand Graph Analytics Keywords: Transit Networks,Mobility Pattern,Time-Varying Graph model, Graph Database and Graph Analytics
Environmental monitoring and management systems in most cases deal with models and spatial analyt... more Environmental monitoring and management systems in most cases deal with models and spatial analytics that involve the integration of in-situ and remote sensor observations. In-situ sensor observations and those gathered by remote sensors are usually provided by different databases and services in real-time dynamic services such as the Geo-Web Services. Thus, data have to be pulled from different databases and transferred over the network before they are fused and processed on the service middleware. This process is very massive and unnecessary communication and work load on the service. Massive work load in large raster downloads from flat-file raster data sources each time a request is made and huge integration and geo-processing work load on the service middleware which could actually be better leveraged at the database level. In this paper, we propose and present a heterogeneous sensor database framework or model for integration, geo-processing and spatial analysis of remote and ...
The interplay between Geographical Information System (GIS) and Computer Science has continued to... more The interplay between Geographical Information System (GIS) and Computer Science has continued to yield improved methods of carrying out many surveying-related activities. In the past, survey control points were stored in file systems and at the best in Database Management applications thereby leading to the limited usage of the survey control points since they are difficult to locate in the field. This study however, suggests another approach for the storage of these survey control points which makes them to be easily accessible and gives room for faster update and geo-visualization of the survey control points. This was achieved by means of web programming applications such as Node-JS, Leaflet Javascript Mapping API, MONGODB, HTML and CSS, integrating GIS into web technologies. The end product is an interactive web application that can be accessed using any smart device with the control points rendered on the user interface. The Survey Control Finder application (E-Beacon) is a We...
Proceedings of the Second ACM/IEEE Symposium on Edge Computing, 2017
Mobility analytics using data generated from the Internet of Mobile Things (IoMT) is facing many ... more Mobility analytics using data generated from the Internet of Mobile Things (IoMT) is facing many challenges which range from the ingestion of data streams coming from a vast number of fog nodes and IoMT devices to avoiding overflowing the cloud with useless massive data streams that can trigger bottlenecks [1]. Managing data flow is becoming an important part of the IoMT because it will dictate in which platform analytical tasks should run in the future. Data flows are usually a sequence of out-of-order tuples with a high data input rate, and mobility analytics requires a real-time flow of data in both directions, from the edge to the cloud, and vice-versa. Before pulling the data streams to the cloud, edge data stream processing is needed for detecting missing, broken, and duplicated tuples in addition to recognize tuples whose arrival time is out of order. Analytical tasks such as data filtering, data cleaning and low-level data contextualization can be executed at the edge of a n...
Fault identification is one of the most significant bottlenecks faced by electricity transmission... more Fault identification is one of the most significant bottlenecks faced by electricity transmission and distribution utilities in developing countries to deliver efficient services to the customers and ensure proper asset audit and management for network optimization and load forecasting. This is due to data scarcity, asset inaccessibility and insecurity, ground-surveys complexity, untimeliness, and general human cost. In view of this, we exploited the use of oblique UAV imagery with a high spatial resolution and a fine-tuned and deep Convolutional Neural Networks (CNNs) to monitor four major Electric power transmission network (EPTN) components. This study explored the capability of the Single Shot Multibox Detector (SSD), a one-stage object detection model on the electric transmission power line imagery to localize, detect and classify faults. The fault considered in this study include the broken insulator plate, missing insulator plate, missing knob, and rusty clamp. Our adapted ne...
Understanding the biases in Deep Neural Networks (DNN) based algorithms is gaining paramount impo... more Understanding the biases in Deep Neural Networks (DNN) based algorithms is gaining paramount importance due to its increased applications on many real-world problems. A known problem of DNN penalizing the underrepresented population could undermine the efficacy of development projects dependent on data produced using DNN-based models. In spite of this, the problems of biases in DNN for Land Use and Land Cover Classification (LULCC) have not been a subject of many studies. In this study, we explore ways to quantify biases in DNN for land use with an example of identifying school buildings in Colombia from satellite imagery. We implement a DNN-based model by fine-tuning an existing, pre-trained model for school building identification. The model achieved overall 84% accuracy. Then, we used socioeconomic covariates to analyze possible biases in the learned representation. The retrained deep neural network was used to extract visual features (embeddings) from satellite image tiles. The ...
Computer vision for large scale building detection can be very challenging in many environments a... more Computer vision for large scale building detection can be very challenging in many environments and settings even with recent advances in deep learning technologies. Even more challenging is modeling to detect the presence of specific buildings (in this case schools) in satellite imagery at a global scale. However, despite the variation in school building structures from rural to urban areas and from country to country, many school buildings have identifiable overhead signatures that make them possible to be detected from high-resolution imagery with modern deep learning techniques. Our hypothesis is that a Deep Convolutional Neural Network (CNN) could be trained for successful mapping of school locations at a regional or global scale from high-resolution satellite imagery. One of the key objectives of this work is to explore the possibility of having a scalable model that can be used to map schools across the globe. In this work, we developed AI-assisted rapid school location mappi...
The study of the dynamic relationship between topological structure of a transit network and the ... more The study of the dynamic relationship between topological structure of a transit network and the mobility patterns of transit vehicles on this network is critical towardsdevising smart and time-aware solutions to transit management and recommendation systems. This paper proposes a time-varying graph (TVG) to model thisrelationship. The effectiveness of this proposed model has been explored by implementing the model in Neo4j graph database using transit feeds generated by bus transit network of the City of Moncton, New Brunswick, Canada. Dynamics in this relationshipalsohave been detected using network metrics such as temporal shortest paths, degree, betweenness and PageRank centralities as well as temporal network diameter and density. Keywords: Transit Networks,Mobility Pattern,Time-Varying Graph model, Graph Databaseand Graph Analytics Keywords: Transit Networks,Mobility Pattern,Time-Varying Graph model, Graph Database and Graph Analytics
Environmental monitoring and management systems in most cases deal with models and spatial analyt... more Environmental monitoring and management systems in most cases deal with models and spatial analytics that involve the integration of in-situ and remote sensor observations. In-situ sensor observations and those gathered by remote sensors are usually provided by different databases and services in real-time dynamic services such as the Geo-Web Services. Thus, data have to be pulled from different databases and transferred over the network before they are fused and processed on the service middleware. This process is very massive and unnecessary communication and work load on the service. Massive work load in large raster downloads from flat-file raster data sources each time a request is made and huge integration and geo-processing work load on the service middleware which could actually be better leveraged at the database level. In this paper, we propose and present a heterogeneous sensor database framework or model for integration, geo-processing and spatial analysis of remote and ...
The interplay between Geographical Information System (GIS) and Computer Science has continued to... more The interplay between Geographical Information System (GIS) and Computer Science has continued to yield improved methods of carrying out many surveying-related activities. In the past, survey control points were stored in file systems and at the best in Database Management applications thereby leading to the limited usage of the survey control points since they are difficult to locate in the field. This study however, suggests another approach for the storage of these survey control points which makes them to be easily accessible and gives room for faster update and geo-visualization of the survey control points. This was achieved by means of web programming applications such as Node-JS, Leaflet Javascript Mapping API, MONGODB, HTML and CSS, integrating GIS into web technologies. The end product is an interactive web application that can be accessed using any smart device with the control points rendered on the user interface. The Survey Control Finder application (E-Beacon) is a We...
Proceedings of the Second ACM/IEEE Symposium on Edge Computing, 2017
Mobility analytics using data generated from the Internet of Mobile Things (IoMT) is facing many ... more Mobility analytics using data generated from the Internet of Mobile Things (IoMT) is facing many challenges which range from the ingestion of data streams coming from a vast number of fog nodes and IoMT devices to avoiding overflowing the cloud with useless massive data streams that can trigger bottlenecks [1]. Managing data flow is becoming an important part of the IoMT because it will dictate in which platform analytical tasks should run in the future. Data flows are usually a sequence of out-of-order tuples with a high data input rate, and mobility analytics requires a real-time flow of data in both directions, from the edge to the cloud, and vice-versa. Before pulling the data streams to the cloud, edge data stream processing is needed for detecting missing, broken, and duplicated tuples in addition to recognize tuples whose arrival time is out of order. Analytical tasks such as data filtering, data cleaning and low-level data contextualization can be executed at the edge of a n...
Uploads
Papers by Ikechukwu Maduako