The edge platform has evolved to become a part of a distributed computing environment. While typical edges do not have enough processing power to train machine learning models in real time, it is common to generate models in the cloud for use on the edge. The pattern of heterogeneous Internet of Things (IoT) data is dependent on individual circumstances. It is not easy to guarantee prediction performance when a monolithic model is used without considering the spatial characteristics of the space generating those data. In this paper, we propose a collaborative framework using a new method to select the best model for the edge from candidate models of cloud based on sample data correlation. This method lets the edge use the most suitable model without any training tasks on the edge side, and it also minimizes privacy issues. We apply the proposed method to predict future fine particulate matter concentration in an individual space. The results suggest that our method can provide better performance than the previous method.
As deep learning technology paves its way, real-world applications that make use of it become popular these days. Edge computing architecture is one of the service architectures to realize the deep learning based service, which makes use of the resources near the data source or client. In Edge computing architecture it becomes important to manage resource usage, and there is research on optimization of deep learning, such as pruning or binarization, which makes deep learning models more lightweight, along with the research for the efficient distribution of workloads on cloud or edge resources. Those are to reduce the workload on edge resources. In this paper, a usage optimization method with batch and model management is proposed. The proposed method is to increase the utilization of GPU resource by modifying the batch size of the input of an inference application. To this end, the inference pipelines are identified to see how the different kinds of resources are used, and then the effect of batch inference on GPU is measured. The proposed method consists of a few modules, including a tool for batch size management which is able to change a batch size with respect to the available resources, and another one for model management which supports on-the-fly update of a model. The proposed methods are implemented on a real-time video analysis application and deployed in the Kubernetes cluster as a Docker container. The result shows that the proposed method can optimize the usage of edge resources for real-time video analysis deep learning applications.
20 December 2024 in Computers and Electronics in Agriculture
With the proliferation of IoT devices, real-time data collection and monitoring services are being offered to various industries. However, time series data, due to its temporal nature, presents numerous unforeseen challenges over time, often leading to data contamination. Such contaminated data make it difficult to effectively utilize in practical applications. This paper investigates the issues with time series data collected from sensors and introduces preprocessing techniques and a corresponding platform to overcome these challenges. In addition to established preprocessing methods, novel approaches for processing and transforming multiple time series data sets are applied, followed by performance testing within the framework of the proposed service platform.
In the industry, various time-series data are being generated, but it is challenging to handle them consistently due to the diverse purposes of data collection and the distinct characteristics of each dataset. Time-series data should be able to be combined and utilized for different purposes beyond their initial storage objectives. However, the development of platforms that search for and utilize a large number of time-series data lags behind other data formats. In this paper, we propose a time-series data management system that stores and manages time-series data for different purposes, facilitates the discovery of relevant time-series data, and enables their combination and processing based on specific objectives. The proposed management platform will assist in effectively finding and utilizing various incomplete and problematic real-time series data. Furthermore, future research is planned to be conducted, leveraging the prepared time-series data in the platform to test various learning and analysis techniques and uncover their underlying meanings.
Due to the development of ICT technology and the explosion of data, various technologies such as smart farms and smart greenhouses have been applied to the agricultural sector, helping farmers a lot. This paper aims to process and utilize crop image datasets collected periodically from a strawberry greenhouse cultivation environment to analyze and monitor crop growth and yields. In addition to the time-series data collected by measuring environmental variables such as temperature, humidity, light intensity, and carbon dioxide concentration, growth-related data are needed to monitor fruit growth and predict yield. Rail cameras were installed on the testbed to collect image data on strawberry every certain time in a specific section, and to construct a time-series image dataset by detecting objects for each stage of strawberry growth. Due to the characteristics of strawberries, we represented the crop images into growth monitoring data applied to the entire fruit of the testbed or crop clusters, not to each fruit. In addition, it can be useful for various use cases if the represented growth-related time series data and sensor data are properly used together.