Adopting the Internet of Things Technology to Remotely Monitor COVID-19 Patients

Since 2019, the world has been living with the pandemic of coronavirus known as COVID-19, and it has impacted many people around the world. COVID-19 has a lot of variants, such as the Delta and Omicron variants, which are characterized by a high propagation speed in a short period of time. Generally, the virus causes respiratory sickness because it attacks the lungs. COVID-19 has three categories of symptoms: most common symptoms such as fever, less common symptoms like loss of taste or smell, and severe symptoms such as difficulty breathing. After much research, some institutes are able to produce vaccines against COVID-19. They are approved by the World Health Organization (WHO). Globally, as of February 18th, 2022, there have been 418,50,474 confirmed cases of COVID-19, including 5,856,224 deaths reported to WHO. As of February 15th, 2022, a total of 10,279,668,555 vaccines have been administered [25]. Despite the vaccines discovered, the battle is not over. Not only because the demand for the purchase of the vaccine is greater than the production and poor countries do not have the financial means to buy it, but also because of the mutations of the virus. For that, we need to reinforce protection by using emerging technologies. The Internet of Things can play an important role in the fight against COVID-19 and can be considered as an enabler by providing smart and innovative solutions. Collecting the data, transmitting data, reacting after the processing of data, and visualizing the data are exactly what the IoT provides to us and what is needed to deal with the virus. IoT, can be applied in different domains like transportation, agriculture, and, in our case, healthcare. It helps to prevent and detect diseases and react at the right time. In the case of COVID-19, we can use an IoT-based wearable body sensor to monitor patients remotely. But, creating applications in this domain is too hard due to the complexity of IoT systems. In addition to the storage constraint, we have a huge number of connected devices, communication, and processing infrastructures. The design of architectural software systems enables architects to grasp the construction of complex software systems. However, their colloquial description may cause confusion, resulting in erroneous software system implementation. In this paper, to address this issue, we propose a meta-model and implement two scenarios using an IoT system to fight COVID-19 using some sensors and actuators in order to test the proposed meta-model. The first one is to monitor contaminated people and alert the concerned authorities. And the second one is to apply social distancing when people wait their turn in supermarkets for payment, in administration, etc.

The remainder of the paper is organized as follows. Section 2 outlines related work. Section 3 presents the proposed meta-model. Section 4 introduces our use cases and their modeling. The necessary equipment, the implementation details, and results are discussed in Sect. 5. The conclusion is reported in Sect. 6.

The Internet of Things refers to the process of connecting the physical world to the Internet, including things such as light bulbs, medical devices, and traffic lights via a network. Process the data collected by sensors in the cloud. The result of the processing phase is actions in order to react and provide the user with useful information through applications. For this, several business sectors can benefit from IoT such as home automation, health, transport, industry and agriculture. Some researchers propose methods and techniques to assist developers and architects from the modeling phase to the final IoT product. Previous works are classified into two categories: those with a high level of abstraction and others with a low level of abstraction. In the second category, we notice different application domains. In the first category, in [5], authors propose a new domain-specific language called BIoT. This makes the creation of software architectures for IoT applications easier for software developers and permits them to avoid errors in the design step. Kallel et al. [9] use a business process model (BPM) and integrate the IoT-fog architecture with the cloud. Heterogeneous and non-heterogeneous IoT resources, resource metrics, and QoS constraints have all been gained by the architecture. Authors in [21] propose a profile called UML4IoT that enables us to exploit IoT in manufacturing systems. It enables the designer to create a cyber-physical component utilizing software and system standards. For the modeling and specification of design patterns, authors in [22, 23] propose an approach to designing the SOA design patterns using the SoaML standard. Next, they attribute to them a formal semantic using the Event-B method. In past work [19], we applied our methodology of modeling IoT systems to a smart home system. In the other category, we find several application domains. In the agriculture domain, authors present an IoT architecture for smart farming [6] based on wireless sensor networks and a plug-and-play approach for standalone nodes. The used algorithm increases the network’s lifetime. In smart home field, paper [12] presents a smart home system that improves the quality of life of a home owner using IoT techniques. The authors implement a server to interconnect things in the home and a web application to control these things in real time. In the healthcare sector, the study [20] designed a system that detects the quality of sleep in a patient by collecting some data with sensors. These are transmitted to a local server in order to use the random forest classification method for predictions and classification. Authors in [13], add a new technology to the IoT, which is Deep Learning. An optimized neural network with an accuracy of 97% can be used in an IoT system to detect fall. Authors in [24], propose an approach that combines IoT, fog, and cloud computing technologies in a healthcare monitoring system for healthy aging in a familiar setting to improve quality of life. In [15], authors predict and analyze indoor air quality, which is important in the case of infected people in quarantine using IoT technology and a machine learning model with high accuracy. Baskaran et al. [3], developed a system that permits the detection of COVID-19 infection in a work environment by detecting if a person is wearing a mask and thermal image. In [8], the authors propose a healthcare system based on the IoT using Neural Networks. The system processes data and makes decisions using the fuzzy logic system. One of the problems caused by COVID-19 is breathing. The lack of oxygen and oxygen concentrators in developed areas prompted the authors in [14] to propose a low-cost device that is easy to build. In the public safety and environmental monitoring domain, papers [1, 2] analyze the air quality in open areas using air quality sensors and air pollutant data. This analytic permit to classify regions if they are good to live in or unhealthy. In [26], the authors propose a framework that integrate a Fog technology. The framework minimized the delay to 8% and had an accuracy rate of 95%. In the transportation field, a previous study [17] proposed a system that would permit monitoring of the platform automatically with the arriving trains using IoT sensors and actuators. Finally, any IoT system needs to be secure (unauthorized persons). For that, authors in [10] propose an intrusion detection system for the smart IoT environment. A system was developed to remotely manage IoT devices using 3G connectivity technology in [16]. We have a vertical layer that across the horizontal layers of IoT which is security layer. For that, a profile called IoTsec is discussed in [18]. This UML/SysML extension claims to describe IoT security knowledge, and it can be considered as a first step to building a robust modelling language for IoT systems in terms of security and hence user safety. In [7] authors give a survey on how to deal with privacy and security while using IoT applications.

By comparing previous works, we find some gabs in the conceptual part as well as the implementation part. However, most of the previous studies do not take into account how the components of an IoT system are interconnected in their models, as well as the nature of the data exchanged in the IoT network. Furthermore, some meta-models or profiles are specific to a particular domain application. For the proposed IoT applications, the major drawback of some applications is that they do not respect the general architecture of the IoT. In addition, they don’t provide the system modeling, which is an important step before the implementation phase. Our work belongs to the two categories mentioned before; we describe our system with a meta-model at a high level of abstraction and we implement it.

We propose a meta-model to describe the Internet of Things architecture as shown in Fig. 1. The proposed meta-model takes into consideration all the components that we can find in an IoT architecture, the connections between these components, and the type of the exchanged messages. The following are the fundamental components of the meta-model:

We have added design patterns to ensure some quality attributes. These design patterns are represented as an interface to be implemented. For example, certain devices can’t connect directly to a network because they don’t support the appropriate communication methods. A DEVICE GATEWAY design pattern proposes a layer in the IoT gateway that translates the communication technology to the right one. Connectivity, cost, and reusability are quality attributes guaranteed with this pattern.

In this section, we present our use cases. Figure 2 shows the general structure of our IoT system. We are dealing with two scenarios. The first one is to capture the corporal temperature (the most common symptom of COVID-19) of contaminated people through a sensor. Also, their heart beat and oxygen saturation. Next, the doctor can visualize the value of the collected data via a website or Android application. If the temperature is high or if there is a breathing problem, the doctor will react immediately by generating an alert in order to avoid any further complications. The second scenario is to apply social distancing in shopping queues or offices in order to avoid the spread of COVID-19 by using sensors and actuators.

In this section, we present the materials used (software and hardware) in the development. A typical IoT architecture is composed of four layers:

In this paper, we presented our proposed meta-model that permits modeling the general architecture of the IoT. We implemented two functional applications to help the world in this pandemic and outbreak the spreading of COVID-19. Firstly, we presented the necessary equipment in both scenarios. Next, we designed correct models by construction using UML diagrams to describe the behavioral and structural views of our system. After that, we implemented the system using the Embedded C (to program the Node MCU) and Java (to program an Android application) languages. Finally, we presented and evaluated our application. The application has shown that it is effective in fighting the coronavirus with the obtained results.

In the short term, we will strengthen our application by adding more sensors to offer more services and actuators to have a good preventive and alerting system.

In the long term, we will integrate new technologies such as artificial intelligence to get the most out of the huge amounts of data that sensors collect. This will allow us to make highly precise diagnoses and reactions.