A smart device for non-invasive ADL estimation through multi-environmental sensor fusion.

This research paper introduces the Smart Plug Hub (SPH), a non-invasive system designed to accurately estimating a patient's Activities of Daily Living (ADL). Traditional methods for measuring ADL include interviews, remote video systems, and wearable devices that track behavior. However, these approaches have limitations, such as patient memory dependency, privacy violations, and careless device management. To address these limitations, SPH utilizes sensor fusion to analyze time-series environmental signals and accurately estimate a patient's ADL. We have effectively optimized the utilization of computing resources through the implementation of "device collaboration" in SPH to receive event data and segments portions of the time-series environmental signal. By segmenting the data into smaller segments, we extracted an analyzable dataset, which was processed by an edge device-SPH. We have conducted several experiments with the SPH, and our research has resulted in a significant 75% accuracy in the classification of patients' kitchen ADLs and an 85% accuracy in the classification of toilet ADLs. These activities include actions such as eating activities in the kitchen and typical activities performed in the toilet. These findings have substantial implications for the progress of healthcare and patient care, highlighting the potential uses of the SPH technology in the monitoring and improvement of daily living activities.

Knowledge-Based Remote E-Coaching Framework Using IoT Devices for In-Home ADL Rehabilitation Treatment of Degenerative Brain Disease Patients.

The activities of daily living (ADL) ability level of an elderly patient is an important indicator in determining the patient's degree of degenerative brain disease and is mainly evaluated through face-to-face interviews with doctors and patients in hospitals. It is impossible to determine the exact ADL ability of a patient through such a temporary interview, and the pursuit of accurate ADL ability evaluation technology is a very important research task worldwide. In this paper, in order to overcome the limitations of the existing ADL evaluation method mentioned above, first of all, a self-organized IoT architecture in which IoT devices autonomously and non-invasively measure a patient's ADL ability within the context of the patient's daily living place was designed and implemented. Second, a remote rehabilitation treatment concept for enhancing the patient's ADL ability we call an "e-coaching framework", in which a doctor remotely gives an instruction in a specific ADL scenario, and the patient's ability to understand and perform the instruction can be measured on-line and in real time, was additionally developed on top of the self-organized IoT architecture. In order to verify the possibility of remote rehabilitation treatment through the proposed architecture, various remotely directed ADL scenarios were performed and the accuracy of the measurements was verified.

Secured and Deterministic Closed-Loop IoT System Architecture for Sensor and Actuator Networks.

Sensors, actuators, and wireless communication technologies have developed significantly. Consequently, closed-loop systems that can be monitored and controlled by devices in IoT environments, such as farms and factories, have emerged. Such systems are realized by means of cloud-level and edge-level implementations. Among them, with a model that generates real-time control decisions at the cloud level, it might be difficult to ensure the determinism of real-time control due to communication overheads. In addition, if the actuator is remotely controlled at the cloud level, it is difficult to secure control safety against external hacking or device malfunction. Herein, we propose a system architecture that can fulfil real-time performance and safety requirements with two commonly used devices, Field Edge Unit (FEU) and Current Sensing Tag (CST), in a closed-loop IoT environment. By using these devices, we designed a special architecture that can be commonly used in various closed-loop sensing and actuating applications. In this study, the proposed architecture is evaluated by applying it to a smart farm application.

Clinical application of the experimental ADL test for patients with cognitive impairment: pilot study.

We employed a hospital-based Internet of Things (IoT) platform to validate the role of real-time activities of daily living (ADL) measurement as a digital biomarker for cognitive impairment in a hospital setting. Observational study. 12 patients with dementia, 11 patients with mild cognitive impairment (MCI), and 15 cognitively normal older adults. The results of 13 experimental ADL tasks were categorized into success or fail. The total number of successful task and the average success proportion of each group was calculated. Time to complete the total tasks was also measured. Patients with dementia, patients with MCI, and cognitively normal older adults performed 13 experimental ADL tasks in a hospital setting. Significant differences in the average success rate of 13 tasks were found among groups. Dementia group showed the lowest success proportion (49.3%) compared with MCI group (78.3%) and normal group (97.4%). Correlation between classical ADL scales and the number of completed ADL tasks was statistically significant. In particular, instrumental ADL (I-ADL) had stronger relationship with the number of completed ADL tasks than Barthel's ADL (B-ADL). Dementia group required more time to accomplish the tasks when compared to MCI and normal groups. This study demonstrated that there is a clear relationship between the performance of experimental ADL tasks and the severity of cognitive impairment. The evaluation of ADLs involving the IoTs platform in an ecological setting allows accurate assessment and quantification of the patient's functional level.

Self-Organizing IoT Device-Based Smart Diagnosing Assistance System for Activities of Daily Living.

Activity of daily living (ADL) is a criterion for evaluating the performance ability of daily life by recognizing various activity events occurring in real life. However, most of the data necessary for ADL evaluation are collected only through observation and questionnaire by the patient or the patient's caregiver. Recently, Internet of Things(IoT) device studies using various environmental sensors are being used for ADL collection and analysis. In this paper, we propose an IoT Device Platform for ADL capability measurement. Wearable devices and stationary devices recognize activity events in real environments and perform user identification through various sensors. The user's ADL data are sent to the network hub for analysis. The proposed IoT platform devices support many sensor devices such as acceleration, flame, temperature, and humidity in order to recognize various activities in real life. In addition, in this paper, using the implemented platform, ADL measurement test was performed on hospital patients. Through this test, the accuracy and reliability of the platform are analyzed.

Real-Time User Identification and Behavior Prediction Based on Foot-Pad Recognition.

In the IoT (Internet of things)-based smart home, the technology for recognizing individual users among family members is very important. Although research in areas such as image recognition, biometrics, and individual wireless devices is very active, these systems suffer from various problems such as the need to follow an intentional procedure or own a specific device. Furthermore, with a centralized server system for IoT service, it is difficult to guarantee real-time determinism with high accuracy. To overcome these problems, we suggest a method of recognizing users in real time from the foot pressure characteristics measured as a user steps on a footpad. The proposed model in this paper uses a preprocessing algorithm to determine and generalize the angle of foot pressure. Based on this generalized foot pressure angle, we extract nine features that can distinguish individual human beings, and employ these features in user-recognition algorithms. Performance evaluation of the model was conducted by combining two preprocessing algorithms used to generalize the angle with four user-recognition algorithms.

A Situation-Aware Indoor Localization (SAIL) System Using a LF and RF Hybrid Approach.

Recently, studies focusing on identifying user's current location for use in a wide variety of differentiated location-based and localization services have steadily increased. In particular, location awareness using wireless communication is gaining attention in indoor environments composed of many obstacles, where GPS signals cannot reach. Previously, studies have focused mostly on location precision, which resulted in using many beacon nodes, not considering the initial installation and maintenance costs, communication robustness, or power consumption. This makes it difficult to apply existing methods to various fields, especially in mobile nodes (i.e., wearable devices, mobile tags, etc.) with limited battery capacity. In this paper, we propose a hybrid situation-aware indoor localization (SAIL) system for real-time indoor localization using a combination of low frequency (LF) and Bluetooth Low Energy (BLE) 4.0. This approach allows us to work with limited battery capacity mobile devices, and identify tagged mobile nodes and their current location in relevance to the anchor node. In our experiment, we attached one anchor node at the entrance to indoor areas such as office or factory settings. Using our hybrid SAIL system, we were able to detect the passing of a mobile node through the entrance and recognize whether the node is entering or exiting the room by calculating the direction of movement as well as the distance from the entrance. This allowed us to distinguish the precise position in an indoor environment with the margin of error being 0.5 m. The signal attenuation due to obstacles is overcome by using LF communication in the 125-kHz band. This approach enables us to reduce the number of initially installed anchor nodes as well as the power consumption of the mobile node. We propose an indoor position recognition system, namely, the hybrid SAIL system, that can be applied to mobile nodes with limited battery capacity by reducing the system complexity and power consumption.

Opportunistic and Location-Based Collaboration Architecture among Mobile Assets and Fixed Manufacturing Processes.

Research into integrating the concept of the internet of things (IoT) into smart factories has accelerated, leading to the emergence of various smart factory solutions. Most ideas, however, focus on the automation and integration of processes in factory, rather than organic cooperation among mobile assets (e.g., the workers and manufactured products) and fixed manufacturing equipment (e.g., press molds, computer numerical controls, painting). Additionally, it is difficult to apply smart factory and IoT designs to analog factories, because such a factory would require the integration of mobile assets and smart manufacturing processes. Thus, existing analog factories remain intact and smart factories are newly constructed. To overcome this disparity and to make analog factories compatible with smart technologies and IoT, we propose the opportunistic and location-based collaboration architecture (OLCA) platform, which allows for smart devices to be attached to workers, products, and facilities to enable the collaboration of location and event information in devices. Using this system, we can monitor workers' positions and production processes in real-time to help prevent dangerous situations and better understand product movement. We evaluate the proposed OLCA platform's performance while using a simple smart factory scenario, thus confirming its suitability.

Self-Organizing Peer-To-Peer Middleware for Healthcare Monitoring in Real-Time.

As the number of elderly persons with chronic illnesses increases, a new public infrastructure for their care is becoming increasingly necessary. In particular, technologies that can monitoring bio-signals in real-time have been receiving significant attention. Currently, most healthcare monitoring services are implemented by wireless carrier through centralized servers. These services are vulnerable to data concentration because all data are sent to a remote server. To solve these problems, we propose self-organizing P2P middleware for healthcare monitoring that enables a real-time multi bio-signal streaming without any central server by connecting the caregiver and care recipient. To verify the performance of the proposed middleware, we evaluated the monitoring service matching time based on a monitoring request. We also confirmed that it is possible to provide an effective monitoring service by evaluating the connectivity between Peer-to-Peer and average jitter.

Middleware Design for Swarm-Driving Robots Accompanying Humans.

Research on robots that accompany humans is being continuously studied. The Pet-Bot provides walking-assistance and object-carrying services without any specific controls through interaction between the robot and the human in real time. However, with Pet-Bot, there is a limit to the number of robots a user can use. If this limit is overcome, the Pet-Bot can provide services in more areas. Therefore, in this study, we propose a swarm-driving middleware design adopting the concept of a swarm, which provides effective parallel movement to allow multiple human-accompanying robots to accomplish a common purpose. The functions of middleware divide into three parts: a sequence manager for swarm process, a messaging manager, and a relative-location identification manager. This middleware processes the sequence of swarm-process of robots in the swarm through message exchanging using radio frequency (RF) communication of an IEEE 802.15.4 MAC protocol and manages an infrared (IR) communication module identifying relative location with IR signal strength. The swarm in this study is composed of the master interacting with the user and the slaves having no interaction with the user. This composition is intended to control the overall swarm in synchronization with the user activity, which is difficult to predict. We evaluate the accuracy of the relative-location estimation using IR communication, the response time of the slaves to a change in user activity, and the time to organize a network according to the number of slaves.

A Self-Organizing Interaction and Synchronization Method between a Wearable Device and Mobile Robot.

In the near future, we can expect to see robots naturally following or going ahead of humans, similar to pet behavior. We call this type of robots "Pet-Bot". To implement this function in a robot, in this paper we introduce a self-organizing interaction and synchronization method between wearable devices and Pet-Bots. First, the Pet-Bot opportunistically identifies its owner without any human intervention, which means that the robot self-identifies the owner's approach on its own. Second, Pet-Bot's activity is synchronized with the owner's behavior. Lastly, the robot frequently encounters uncertain situations (e.g., when the robot goes ahead of the owner but meets a situation where it cannot make a decision, or the owner wants to stop the Pet-Bot synchronization mode to relax). In this case, we have adopted a gesture recognition function that uses a 3-D accelerometer in the wearable device. In order to achieve the interaction and synchronization in real-time, we use two wireless communication protocols: 125 kHz low-frequency (LF) and 2.4 GHz Bluetooth low energy (BLE). We conducted experiments using a prototype Pet-Bot and wearable devices to verify their motion recognition of and synchronization with humans in real-time. The results showed a guaranteed level of accuracy of at least 94%. A trajectory test was also performed to demonstrate the robot's control performance when following or leading a human in real-time.

Self-organizing distributed architecture supporting dynamic space expanding and reducing in indoor LBS environment.

Indoor location-based services (iLBS) are extremely dynamic and changeable, and include numerous resources and mobile devices. In particular, the network infrastructure requires support for high scalability in the indoor environment, and various resource lookups are requested concurrently and frequently from several locations based on the dynamic network environment. A traditional map-based centralized approach for iLBSs has several disadvantages: it requires global knowledge to maintain a complete geographic indoor map; the central server is a single point of failure; it can also cause low scalability and traffic congestion; and it is hard to adapt to a change of service area in real time. This paper proposes a self-organizing and fully distributed platform for iLBSs. The proposed self-organizing distributed platform provides a dynamic reconfiguration of locality accuracy and service coverage by expanding and contracting dynamically. In order to verify the suggested platform, scalability performance according to the number of inserted or deleted nodes composing the dynamic infrastructure was evaluated through a simulation similar to the real environment.

Fully distributed monitoring architecture supporting multiple trackees and trackers in indoor mobile asset management application.

A tracking service like asset management is essential in a dynamic hospital environment consisting of numerous mobile assets (e.g., wheelchairs or infusion pumps) that are continuously relocated throughout a hospital. The tracking service is accomplished based on the key technologies of an indoor location-based service (LBS), such as locating and monitoring multiple mobile targets inside a building in real time. An indoor LBS such as a tracking service entails numerous resource lookups being requested concurrently and frequently from several locations, as well as a network infrastructure requiring support for high scalability in indoor environments. A traditional centralized architecture needs to maintain a geographic map of the entire building or complex in its central server, which can cause low scalability and traffic congestion. This paper presents a self-organizing and fully distributed indoor mobile asset management (MAM) platform, and proposes an architecture for multiple trackees (such as mobile assets) and trackers based on the proposed distributed platform in real time. In order to verify the suggested platform, scalability performance according to increases in the number of concurrent lookups was evaluated in a real test bed. Tracking latency and traffic load ratio in the proposed tracking architecture was also evaluated.

A mobile asset tracking system architecture under mobile-stationary co-existing WSNs.

The tracking of multiple wireless mobile nodes is not easy with current legacy WSN technologies, due to their inherent technical complexity, especially when heavy traffic and frequent movement of mobile nodes are encountered. To enable mobile asset tracking under these legacy WSN systems, it is necessary to design a specific system architecture that can manage numerous mobile nodes attached to mobile assets. In this paper, we present a practical system architecture including a communication protocol, a three-tier network, and server-side middleware for mobile asset tracking in legacy WSNs consisting of mobile-stationary co-existing infrastructures, and we prove the functionality of this architecture through careful evaluation in a test bed. Evaluation was carried out in a microwave anechoic chamber as well as on a straight road near our office. We evaluated communication mobility performance between mobile and stationary nodes, location-awareness performance, system stability under numerous mobile node conditions, and the successful packet transfer rate according to the speed of the mobile nodes. The results indicate that the proposed architecture is sufficiently robust for application in realistic mobile asset tracking services that require a large number of mobile nodes.