From Offline to Real-Time Distributed Activity Recognition in Wireless Sensor Networks for Healthcare: A Review

Abstract

This review presents the state of the art and a global overview of research challenges of real-time distributed activity recognition in the field of healthcare. Offline activity recognition is discussed as a starting point to establish the useful concepts of the field, such as sensor types, activity labeling and feature extraction, outlier detection, and machine learning. New challenges and obstacles brought on by real-time centralized activity recognition such as communication, real-time activity labeling, cloud and local approaches, and real-time machine learning in a streaming context are then discussed. Finally, real-time distributed activity recognition is covered through existing implementations in the scientific literature, and six main angles of optimization are defined: Processing, memory, communication, energy, time, and accuracy. This survey is addressed to any reader interested in the development of distributed artificial intelligence as well activity recognition, regardless of their level of expertise.

Soft Fiber Electronics Based on Semiconducting Polymer

“…23,118,119 Second, they can be used for smart medicine and daily healthcare. Currently, medical treatment and health monitoring are primarily dependent on centralized medical institutions, 120 which are unable to perform real-time health-care monitoring and timely treatment of patients. Although some portable electronic products such as smart wristbands and watches could help to some extent, more powerful devices with better wearability (flexible, comfortable, accurate, stable, and low-cost) for realtime health monitoring are still required.…”

“…Employed in metaverse applications, semiconducting polymer-based smart textiles could produce “immersion” in virtual reality a seamless transition between the virtual world and the real world by performing functions such as sensing, communication, computing, and display. ,, Second, they can be used for smart medicine and daily healthcare. Currently, medical treatment and health monitoring are primarily dependent on centralized medical institutions, which are unable to perform real-time health-care monitoring and timely treatment of patients. Although some portable electronic products such as smart wristbands and watches could help to some extent, more powerful devices with better wearability (flexible, comfortable, accurate, stable, and low-cost) for real-time health monitoring are still required.…”

Cloud-Based Internet of Things in Healthcare Applications: A Systematic Literature Review

“…These sensors capture physical parameters of the environment that can be deployed in any place. Also, wearable sensors are the chief types of sensors integrated into wearable objects such as accelerometers, gyroscopes, magnetometers, and RFID readers [111]. The potential of RFID tag sensors has been studied in different industrial sectors including healthcare, food safety, environmental pollution, anti-counterfeiting of banknotes and fake medicines, factories, customer shopping behavior, logistics, public transport, and safety.…”

“…Results of the Liu et al [61] study have achieved 96.92% accuracy in recognizing the patients' activities (e.g., coughing, walking, standing, sitting, and lying) by a wearable ECG monitoring device embedded with accelerometer sensors. In this regard, Baghezza et al [111] have proposed that the combination of environmental sensors (audio, temperature, humidity, and light) with wearable sensors leads to multi-user activity recognition. For example, a wrist-worn accelerometer and environmental sensors lead to better movement modeling that is appropriate for physical therapy applications.…”

Investigation on eXtreme Gradient Boosting for cutting force prediction in milling

“…The offline method is usually used for its fast computation time, however, it does not present the advantage that the online method has to train the model on the go. Therefore, today, it is difficult to use the offline for real-time prediction (Baghezza et al, 2021;Tercan & Meisen, 2022). The research conducted by Traini et al in 2019 presented a framework that is based on a similar extraction method, using the milling data.…”