Evaluation of IoT-Based Smart Home Assistance for Elderly People Using Robot

Abstract

In the development of Internet-of-things (IoT)-based technology, there is a pre-programmed robot called Cyborg which is used for assisting elderly people. It moves around the home and observes the surrounding conditions. The Cyborg is developed and used in the smart home system. The features of a smart home system with IoT technology include temperature control, lighting control, surveillance, security, smart electricity, and water sensors. Nowadays, elderly people may not be able to maintain their homes appropriately and may feel uncomfortable performing day-to-day activities. Therefore, Cyborg can be used to carry out the activities of elderly people. Such activities include switching off unnecessary lights, watering plants, gas control, monitoring intruders or unknown persons, alerting elderly people in emergency situations, etc. These activities are controlled by web-based platforms as well as smartphone applications. The issues with the existing algorithms include that they are inefficient, require a long time for implementation, and have high storage space requirements. This paper proposes the k-nearest neighbors (KNN) with an artificial bee colony (ABC) algorithm (KNN-ABC). In this proposed work, KNN-ABC is used with wireless sensor devices for perceiving the surroundings of the smart home. This work implements the automatic control of electronic appliances, alert signal processors, intruder detection, and performs day-to-day activities automatically in an efficient way. GNB for intruder detection in the smart home environment system using the Cyborg human intervention robot achieved an accuracy rate of 88.12%, the Artificial Bee Colony algorithm (ABC) achieved 90.12% accuracy on the task of power saving in smart home electronic appliances, the KNN technique achieved 91.45% accuracy on the task of providing a safer pace to the elderly in the smart home environment system, and our proposed KNN-ABC system achieved 93.72%.

1. Introduction

Due to the development of technology, dynamic changes have occurred in the automation and application of robotics and related systems. Nowadays, robotics plays a vital role in various applications, reducing the workload of human beings as well as errors made by humans. Robots are used in different surveillance processes such as detection of gas leaks and minimizing the risk of disaster through leakage in the chemical industry. Surveillance is the process of closely monitoring an industry, person, or group in the same and different situations. Surveillance is mainly needed in monitoring public places, border areas, companies, and industries in which the intervention of humans is difficult. This surveillance takes place with the help of an embedded system of robots. A robot is a pre-programmed electronic machine that replaces human work through automation and provides accurate results while minimizing error and improving time efficiency [1]. IoT-based devices are linked with one another by a network that connects electronic home appliances, vehicle-based electronic devices, actuators, and software, allows the exchange of information between one device and another. IoT devices can interact with other devices via Wi-Fi communication module by using the wireless sensor networks (WSN) in smart home electronic appliances and by Low Power Wireless Personal Area Networks (LoWPAN) using RFID (Radio-Frequency Identification). An IoT-based smart home environment operates sensor-based devices remotely using mobile applications [2,3].

Human Interaction Robots (HIR) are mainly used in activities with a social component, such as medicine, neuroscience, cognitive science, and robotics. In order to provide security, the need for human intervention can be replaced with Cyborg. This robot can assist elderly people who are home alone, helping them to avoid crime due to home invasion or theft. In this case, it is necessary to provide security to elderly people by implementing a smart home environment system that contains the required sensor devices and it can transmit sensor signals through a communication module in order to alert the user and allow them to take precautionary steps [4,5]. The smart home secure environment enhances the lifestyle of human beings by providing security, detecting gas leakage in the kitchen, monitoring temperature and humidity in the home, detecting intruders, and more. This can be achieved by monitoring the surroundings of the smart home using a Raspberry Pi-based wireless camera, capturing images with related information, and sending it to the server. The main components of Cyborg are DC motors, a battery, and a wheel chassis, and it can be implemented in either automatic or manual mode [6].

Many research works have been implemented in smart home environments. The main issues are that they are inaccurate and inefficient, consumption time is high, and large amounts of storage space are required. This paper proposes a smart home environment for assisting elderly people using the KNN-ABC technique. It uses sensor-based electronic home appliances to monitor the surroundings of the smart home, detect intruders, and generate an alert notification to a registered mobile device or through a mobile app.

The contributions of this work are as follows:

• To implement smart home assistance for elderly people by using the KNN algorithm to monitor the status of electronic home appliances and provide an ON/OFF state using Cyborg.
• To save energy in smart home systems using the Artificial Bee Colony (ABC) algorithm.
• Analysis of the proposed KNN-ABC using the metric measures of precision, recall, F1-score, and accuracy.

This paper is written in five sections including this introduction. In the remainder of this paper, Section 2 discusses relevant previous works on smart home systems, Section 3 describes the proposed methodology, Section 4 describes the results and evaluates the outcomes, and Section 5 concludes the paper.

2. Related Work

Smart home electronic appliances based on IoT technology require automatic ON/OFF operation using a remote control-based application, voice-based technology, or fixed-time scheduling. A notification can be sent to the user by the server. This control is completely based on the activities of the user and passing the commands which can be triggered the activities through the mobile phone [7,8,9]. C. Victor et al. [10] proposed an IoT-based sensor system for monitoring the temperature in the environment. Using a temperature sensor, the system can collect sensor signals and store them in the server. Gladence et al. [11] proposed a client–server-based machine learning algorithm implemented for establishing an automated smart home environment control system able to interact with humans who send commands or triggering the smart appliances. M. Wendy et al. [12] presented a review of effective smart home technology to support elderly people in aspects related to health and security issues. Mehmood et al. [13] proposed an innovative concept involving managing a cloud storage platform, detecting hindrances, activating IoT devices by passing commands, executing those commands, and then transmitting the information to the registered users via mobile notification. To monitor health-related issues for elderly people in smart homes, various machine learning algorithms (LSTM, SVM, and RNN) can be used. IoT devices can closely observe health conditions of elderly people, analyze their symptoms, and make predictions related to disease, as well as helping patients to consult their physicians and alert them to take medicine at the proper time [14]. Sensor devices are used with wireless networks, software, and computers to detect threats which affect the smart home environment. The implementation of the CNN model produces efficient detection of threats [15]. The Cyborg system can be used to save power, as it is able to automatically switch unnecessary electronic devices into the OFF state. In addition, it can detect the presence of human beings in the external surroundings of the smart home. At the same time, it can send a notification to the resident to perform important activities such as taking medicine, watering plants, etc. The proposed smart home system interfaces with sensor devices and assists elderly people in the smart home environment based on the generated sensor signals [16].

Table 1. List of related works on smart home environment systems and on the technologies and sensor measurement approaches employed by the respective systems.

Author	Technology	Monitoring Function
Ruili Zheng (2022) [17]	IoT	Indoor smart monitoring and modern lifestyle.
X. Gengyi (2021) [18]	Machine Learning algorithm SVM	Energy monitoring in a smart home system
C. Zhou et al. (2021) [19]	Virtual Reality	Classifying human activity with R&D
P. Sharma et al. (2020) [20]	Cloud server-based NodeMCU	Electricity measurements in a smart home
O. Taiwo et al. (2020) [21]	Zigbee, Bluetooth, and Arduino technologies	Health care monitoring system
M. S. Soliman et al. (2020) [22]	LabVIEW	PIR Motion Sensor, indoor detection in a smart home
M. Naing et al. (2019) [23]	Arduino-based smart home control	Temperature, humidity sensor
R. D. Manu et al. (2019) [24]	IoT-based Deep Learning algorithm	Motion sensor, PIR sensor
S. K. Saravanan et al. (2019) [25]	Android-based smart home control	Multiple authentication processes, privacy preservation using key generation
L. D. Liao et al. (2019) [26]	Android-based smart home control	health care monitoring, multifunctional operating systems
D. Popa et al. (2019) [27]	Deep Learning	Energy reduction and power saving

enumerates related works on smart home environment systems along with the technology and sensor measurements employed by the respective systems.

Many earlier works demonstrated the use of IoT technology for energy efficiency, monitoring, and activity detection in a smart home environment. Below, we present selected works, which are tabulated in Table 1 along with their prominent features.

In [17], the author presented a smart home remote control system based on wireless sensor networks that collect positioning information and use actuators to control electrical appliances and operate alarms. In [18], X. Gengyi applied support vector machine (SVM) in a smart-home energy monitoring system using a cloud computing-based platform. The proposed solution improves energy efficiency and makes it easier for human interaction. In [19], C. Zhou et al. proposed a design for a smart home system based on virtual reality. Virtual reality was used to improve control interaction in the smart home. Their experimental results indicated that control methods could be simplified and costs reduced by as much as twenty percent through the use of virtual reality. In [20], P. Sharma et al. proposed a design for an IoT system using NodeMCU for real-time supervision of sensor measurements, allowing the user to control electrical loads in a smart home. O. Taiwo et al. [21] proposed a smart home automation mobile application that uses an Arduino microcontroller and personal area communication technologies such as Zigbee and Bluetooth. The practicality of the system was demonstrated through a simulation of the smart home environment.

In [22], M. S. Soliman et al. proposed a smart home automation system based on Arduino and Labview that allows the user to control temperature, save energy, and detect intruders. M. Naing et al. (2019) [23] demonstrated a proposed smart home automation system through a prototype implementation employing two Arduino Nano sensors. Sensors for measuring temperature, smoke, and motion were interfaced with these microcontrollers, which in turn interfaced with actuators to control and secure the home. R. D. Manu et al. (2019) [24] proposed a smart home system able to measure and respond to human activities using long-short term memory (LSTM) deep learning-based decision-making. S. K. Saravanan et al. (2019) [25] proposed a smart home controller using Arduino and Android. A smart door actuator was secured using a multi-factor authentication mechanism. L. D. Liao et al. (2019) [26] proposed the design of a smart home system using Arduino–Uno that provides user control and monitoring through a mobile application. Temperature and motion sensors were connected and controlled by the system to demonstrate its application in a smart home environment.

D. Popa et al. (2019) [27] demonstrate a smart home application where measurements of energy consumption and other sensor data could be stored on a cloud and later analyzed using machine learning methods for improved environmental sustainability and energy efficiency.

The authors of [28] applied linear discriminant analysis to classify power quality disturbances and carry out a performance analysis using KNN, naive Bayes, support vector machine (SVM), and random forest (RF) classifiers. Their results showed that higher classification accuracy was obtained in the presence of noise. In [29], Moraes et al. used a naive Bayes algorithm to propose a structured data mining model that can predict whether a smaller enterprise can join a business association with given attributes. The proposed approach can be utilized as a decision assistance tool for business associations to choose member enterprises. In [30], the authors used four different classifiers, i.e., KNN, naive Bayes, decision tree, and random forest approaches, to distinguish between defective and non-defective metal parts using laser-induced-breakdown spectroscopy. The above-mentioned works show that machine learning algorithms can be used to make accurate predictions and to inform decisions in many situations and for a variety of data formats.

To make the literature survey more comprehensive, below we include several recent optimization methods for feature selection and classification. The authors of [31] proposed a hybrid feature selection method using a combination of the Butterfly optimization algorithm and the Ant Lion optimizer for breast cancer prediction. The proposed hybrid method outperforms both component methods for breast cancer diagnosis in terms of accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve.

Chakraborty et al. [32] proposed an improved whale optimization method for segmentation of chest X-Rays from patients with symptoms of COVID-19. During the global search phase, a random initialization is used to exploit after exploration. The proposed method outperformed the original method in terms of segmentation accuracy.

Sayed et al. [33] adopted a hybrid approach combining a convolutional neural network with Bald Eagle optimization to improve detection performance in melanoma skin cancer prediction. The robustness and accuracy of the proposed approach were verified as being superior through a comparison with state-of-the-art methods.

Xing et al. [34] proposed a modified whale optimization method using a quasi-Gaussian “bare bones” method. The modified method was able to promote diversity and expand the scope of the solution space.

Piri et al. [35] proposed a modified optimization method based on the Harris Hawk optimizer. This method, called multi-objective quadratic binary Harris Hawk optimization, uses a KNN classifier to extract the optimal feature subsets. The proposed methodology proved superior thanks to its better combination of fitness assessment criteria.

3. Proposed Methodology

This section describes the proposed IoT-based Smart Home Assistance system for elderly people using Cyborg.

3.1. Cyborg in Smart Home Assistance Technology

The proposed model KNN-ABC with the Cyborg is designed for applications in monitoring the surroundings of the smart home and assisting elderly people. This robotic model is specifically designed to switch off unnecessary lights, water plants, detect gas leaks, monitor for intruders or unknown persons, alert elderly residents in emergency situations, etc. These tasks are handled by sensor devices connected via Wi-Fi and control interfaces. This control system interface in the smart home allows elderly people to easily access activities in an efficient manner. It satisfies the basic requirements of elderly people by cleaning the home, detecting gas leaks in the kitchen, setting alarms and reminders about important work, and more. The KNN-ABC system uses PIR (Passive Infrared Sensor), a type of electronic sensor used to set alerts for security and automatic ON/OFF operation of fans and lights. The presence or absence of a human in the room is detected using the ZIGBEE communication protocol. For the detection of gas leaks, an LPG sensor is used to produce an alert signal. To clean the house using a vacuum cleaner, the Cyborg robot is moved around the smart home surroundings using the follower technique. The IR sensor (Infrared Sensor) is used to detect obstacles along the movement path of the robot, and it produces the buzzer that makes a beeping sound. For reminders about medication scheduling and important activities, a real-time clock (RTC) is used to set alarms and produce reminder messages on the LCD screen. Figure 1 shows the overall diagram of the proposed smart home assistance system for elderly people using KNN-ABC.

Figure 1 shows the smart home assistance control for elderly people that controls the electronic appliances, sensor devices, electronic home appliances, control system interfaces, and cloud-based computing platform modules. The KNN-ABC system module contains the home environment, control system interface for the user, sensor-based electronic home appliances, and cloud computing platform. The user can communicate with the home environment via Wi-Fi using either a mobile-based or web-based application. The home environment contains sensor devices and communication modules, and can be linked with electronic home appliances. The received sensor signals and their information are stored in the cloud storage platform back-end of the KNN-ABC system model.

3.1.1. Control System Interface

In the KNN-ABC model, the interaction between the user and the home environment at the front-end takes place in the control system interface via Wi-Fi through the mobile-based or web-based application. The surveillance camera live stream monitors the surroundings of the home environment. Using the customized Android application, users can easily access the smart home surroundings status (e.g., humidity, temperature, ON/OFF status of electronic home appliances such as fans and lights, presence of intruders). Data generated by the various sensors are stored in the cloud computing platform for future reference. The control application of the smart home is uses Android. The live stream of the surveillance camera can be displayed on a laptop, mobile device, or desktop computer.

3.1.2. Smart Home Environment

The KNN-ABC-based smart home environment comprises three modules: the communication interface module, electronic home appliances module, and sensor device module. The sensor devices are incorporated with IoT-based electronic home appliances and linked together using a microcontroller to provide communication with the smart home’s outer surrounding environment via the wireless network. The microcontroller uses the ESP8266 module and Wi-Fi with HTTPS/IP and TCP/IP as the communication protocol. In order to function, the microcontroller requires a power supply and the ESP32 camera module in an Arduino board. The in-built function of the Wi-Fi chip has an ESP32 camera board with wireless connectivity. Therefore, ESP8266 and ESP32 are used for communication.

3.1.3. Cloud Computing Platform

This is an intelligent module that sends commands to the system to carry out various activities on behalf of elderly residents. It comprises a cloud storage system that stores the signals received from sensor devices for future reference. The smart home assistance system for elderly people is enhanced using the KNN-ABC algorithm, and is used for monitoring the surroundings of the smart home environment, detecting gas leaks, monitoring the surveillance camera, automatically turning electronic home appliances on and off, generating alert signals for medication scheduling, and providing other alert notifications about any unusual activities that may be occurring inside the smart home environment. The steps involved in the Smart Home Assistance System for Elderly People using Cyborg is shown in Figure 2.

3.2. K-Nearest Neighbour Algorithm (KNN)

The smart home environment uses a robot named Cyborg, which is used to assist elderly people by switching off unnecessary lights, watering plants, gas control, monitoring for intruders, providing alerts in emergency situations, etc. Sensor signals are collected from various electronic home appliances

$$\begin{array}{c}\hfill s=\{e{a}_1,e{a}_2,e{a}_3,\dots ,e{a}_n\},\end{array}$$

(1)

where s represents the sensor signal values and $e{a}_1,e{a}_2,e{a}_3,\dots ,e{a}_n$ represent the various electronic home appliances.

The sensor signals of electronic home appliances are used to detect the ON/OFF states of appliances perform other functions such as intruder detection using the KNN algorithm. If an intruder is detected, the sensor signal values of electronic home appliances and coordinate value of the safer place are collected and stored in the system as a dataset. When the KNN-ABC system model is activated, it analyzes and detects the nearest coordinate value from the stored is dataset and predicts a safe place for the residents to go. The steps involved in the implementation of KNN algorithm are provided below:

• KNN-1: Used to train the model of KNN-ABC model using dataset.
• KNN-2: Used to evaluate the distance between the user’s location and the coordinate position of value in the dataset, as follows:

  $$\begin{array}{c}\hfill dist(p_1,p_2)=\sqrt{\sum _{i=1}^n{\left(p{1}_i-p{1}_t\right)}^2+{\left(p{2}_i-p{2}_t\right)}^2}\end{array}$$

  (2)
  Here, $dist(p1,p2)$ is the Euclidean distance between the current location and the targeted location in the dataset.
• KNN-3: Sorts all distances in ascending order to select the nearest point.
• KNN-4: The value of $dist(p1,p2)$ is processed by the ABC algorithm as a control input for ON/OFF control of electronic home appliances in the smart home system. It can monitor the sensor signal values using a mobile-based or web-based application.

3.3. Artificial Bee Colony Algorithm for Power Saving in Smart Home System

In the smart home, the environment uses a robot named Cyborg to assist elderly people by switching off unnecessary lights, watering plants, gas control, monitoring for intruders, providing alerts during emergency situations, etc. Using the ABC algorithm, it detects the nearest safe place for residents to move to during an emergency situation. Similarly, the collected sensor signals from the electronic home appliances are stored in the dataset. to obtain more accurate and efficient detection of the ON/OFF state, the Artificial Bee Colony algorithm (ABC) is implemented. The ABC algorithm functions by connecting a socket system with the camera sensor and reading the sensor signals from the various electronic home appliances, gas sensor, camera, and PoseNet human positional sensor. It collects all real time information, including the position of human beings in the smart home, at regular intervals of time and stores it in the cloud storage platform in a dataset using the communication module. The locations of human beings can then be retrieved from cloud storage and compared with the location of electronic appliances to check the distance between the human being and the electronic appliances. If the human being is far away from electronic appliances, then the system receives instructions to turn off the unnecessary smart electronic appliances; otherwise, it can use the information to operate the smart electronic appliances in safety mode. The ABC procedure is explained below.

• ABC 1: Initialize the population size; M is the initial nectar coordinate value and maximum number of iterations.
• ABC 2: Search for a new nectar source by selecting bees from the total number of number of bees N; the bee group size is M, and the spatial dimensionality for bees searching for new nectar sources is S. From the current nectar source, a bee starts its searching process within its neighborhood. The newly created nectar source contains:

  –

  ABC 2.1: The spatial dimension space of the current nectar source is split into regular intervals based on the following formula:

  $$\begin{array}{c}\hfill R_{k,l}^h=Q_k^l+\frac{(2h-H)}{H}(Q_k^l-Q_n^l),h\in \left[0,H\right],\end{array}$$

  (3)

  where R denotes the h-th interval of point from division of the current nectar source, Q denotes k-th current honey source generated in the l-th dimension space, and Q denotes the n-th current honey source generated in the l-th dimension space.

  –

  ABC 2.2: In each interval of R, the interval is divided into several sub-intervals Y based on the formula

  $$\begin{array}{c}\hfill R_{k,l}^{h,z}=R_{k,l}^h+sin\left(\frac{y-rand(0,1)}{2y}\pi \right)\left(R_{k,l}^h-R_{k,l}^{h+1}\right),y\in \left[1,Y\right]\end{array}$$

  (4)

  where R denotes the y-th sub-interval of the current nectar source, $rand(0,1)$ denotes the random distribution of values between 0 and 1 at a uniform rate, and $R+1$ denotes the $R+1$-th interval point produced by division of the next current nectar source.

  –

  ABC 2.3: For every sub-interval of the current nectar source, its fitness function is calculated, then the sub-interval point is selected based on the largest fitness value.

  –

  ABC 2.4: The difference between Q and the fitness value representing the nectar source is evaluated using the following formula:

  $$\begin{array}{c}\hfill F_k^l=min\{fit\left(F_{k,l}^h\right)-fit\left(R_k^l\right)\},fit\left(F_{k,l}^h\right)-fit\left(R_k^l\right)>0,h\in [1,H]\end{array}$$

  (5)

  where $F_k$ denotes the difference between Q and the fitness value of the nectar source, $fit\left(v\right)$ represents the fitness value of thee nectar source in the regular interval of R, and $fit\left(q\right)$ denotes the fitness value of R.

  –

  ABC 2.5: $F_i$ is selected as the nectar source, and it is treated as a new source of nectar.

• ABC 3: Compute the fitness value for newly created nectar source compared to the current nectar source.
• ABC 4: Compare the fitness value of the newly created nectar source with the current nectar source.
• ABC 5: Discard the nectar source with the lower fitness value.
• ABC 6: Based on the probability of the nectar source value, select pickers for following the bees.
• ABC 7: For the selected pickers, update the nectar source values.
• ABC 8: For the current nectar source value, search the closest nectar source.
• ABC 9: Repeat steps 2–5 and retain the nectar source with the largest fitness value.
• ABC 10: Increment the iteration.
• ABC 11: Stop the search process when it reaches the maximum iteration and choose the highest fitness value as the coordinate value of the target node in the nectar source.

The ABC algorithm can achieve the an effective search process in terms of locating the target node; it has good accuracy for determining the position of human beings in a smart home environment and can provide power savings for electronic home appliances in the smart home control system. Despite these advantages, it is inefficient in remote control of smart electronic home appliances. Therefore, the ABC algorithm is modified by implementing the following steps.

• KNN-ABC 1: Randomly generate the initial nectar source from the values of M based on the target nectar source of the bees. Based on the target nectar source with the maximum fitness value, randomly generate the initial nectar source values N using

  $$\begin{array}{c}\hfill Q_l^{l0}=Q_{min}^{l0}+rand(0,1)(Q_{max}^{l0}-Q_{min}^{l0}),\end{array}$$

  (6)

  where $Q_l$ is the l-th initial honey nectar source generated in the k-th dimension spatial space, $Q_{min}$ is the minimum nectar source value of the k-th source of honey, $Q_{max}$ represents the k-th source of honey generated in the spatial dimensional space, and $rand(0,1)$ generates random numbers between 0 and 1 and is uniformly distributed in the system.
• KNN-ABC 2: For every nectar source of honey, compute the reverse honey nectar source using

  $$\begin{array}{c}\hfill Q_l^{l{0}^{\prime }}=rand(0,1)(Q_{max}^{l0}-Q_{min}^{l0})-Q_l^{l0},\end{array}$$

  (7)

  where $Q_l$ is the the reverse nectar source of honey for the l-th initial generation of honey in the k-th spatial dimensional space.
• KNN-ABC 3: Compute the fitness values for all initial nectar sources as well as for the reverse nectar sources. Based on their fitness values, arrange them in descending order to generate the nectar source value set. The first N nectar honey sources are chosen as the target nodes.
• KNN-ABC 4: To improve the quality of the initial nectar honey source by ensuring an even distribution of nodes in the system, in the smart home environment control system the position of the target node is based on the terms of its speed and stability. This allows for more accurate remote control of all smart electronic home appliances. The fitness value is evaluated using

  $$\begin{array}{c}\hfill fit\left(F\right)=min\left(\sqrt{\left({(a-q_l)}^2+{(b-q_l)}^2-D{\lambda }_{pi}\right)}\right),k=1,2,\dots ,\varphi \end{array}$$

  (8)

  where $fit\left(F\right)$ denotes the fitness value of the nectar honey source F at the position of the l-th beacon node, D denotes the average hop distance which is sent by the first beacon node of nectar Q, $q_l$ denotes the total number of hops between the l-th beacon node and the source of honey, and $\varphi$ is the number of beacon nodes.

Using Equation (8) and the modified artificial bee colony algorithm (ABC) improves the accuracy of finding the target node while minimizing the error rate, and is able to provide both effective remote control of smart home electronic appliances and position monitoring of the electronic home appliances via exact angle measurement.

4. Result and Discussion

The proposed KNN-ABC system was implemented in Python 3.6, and was compared with Gaussian Naive Bayes (GNB) [36], Artificial Bee Colony algorithm (ABC) [17], and KNN [37].

Table 2. Metrics of Precision and Recall in Smart Home System.

Algorithm	Precision	Recall	Sensitivity
GNB	65.16%	55.35%	68.76%
ABC	72.11%	67.65%	70.37%
KNN	66.78%	68.46%	71.11%
KNN-ABC	88.32%	78.54%	83.65%

shows the measures of Precision, Recall and Sensitivity for the different algorithms used in the smart home system. Data were collected from the Kaggle website [38].

Precision: Precision quantifies the number of true positive predictions provided by a given technique. It is calculated as follows:

$$\begin{array}{c}\hfill \mathrm{Precision}=\frac{TP}{TP+FP}\times 100.\end{array}$$

(9)

Recall: The percentage of correctly classified true positive predictions is evaluated by calculating the recall, as follows:

$$\begin{array}{c}\hfill \mathrm{Recall}=\frac{TP}{TP+FN}.\end{array}$$

(10)

F1-Score: The F1-Score is a measure of accuracy based on precision and recall values. It is calculated as follows:

$$\begin{array}{c}\hfill \mathrm{F}1-\mathrm{Score}=\frac{2\times \mathrm{Precision}\times \mathrm{Recall}}{\mathrm{Precision}+\mathrm{Recall}}.\end{array}$$

(11)

Specificity: Specificity is used to measure the proportion of actual negative cases that a technique rightly predicts. Specificity is calculated as follows:

$$\begin{array}{c}\hfill \mathrm{Specificity}=\frac{TN}{TN+FP}\times 100.\end{array}$$

(12)

Accuracy:

$$\begin{array}{c}\hfill \mathrm{Accuracy}=\frac{TP+TN}{TP+TN+FP+FN}\times 100.\end{array}$$

(13)

MSE: The mean squared error (MSE) calculates the average of the squares of the differences between the predicted values and actual values.

$$\begin{array}{c}\hfill \mathrm{MSE}=\frac{1}{n}\sum _{i=1}^n{(y_{pi}-y_{ai})}^2.\end{array}$$

(14)

MAE: The mean absolute error (MAE) calculates the average of the squares of the differences between the predicted values and actual values.

$$\begin{array}{c}\hfill \mathrm{MAE}=\frac{1}{n}\sum _{i=1}^n\left|y_{pi}-y_{ai}\right|.\end{array}$$

(15)

Table 2 shows a performance comparison between the proposed KNN-ABC technique and existing algorithms. Here, GNB using Smart Home assistance for elderly people using the Cyborg robot reached a sensitivity of 68.76%, precision of 65.16%, and recall of 55.35%, the Artificial Bee Colony algorithm (ABC) for power saving in smart home electronic appliances reached a sensitivity of 70.37%, precision of 72.11%, and recall of 67.65%, and the KNN technique for detecting the a nearest safer place in an emergency situation reached a sensitivity of 71.11%, precision of 66.78%, and recall of 68.46%. The proposed KNN-ABC technique for detecting intruders, finding the nearest safe place in an emergency, and saving power on smart home electronic appliances attained a sensitivity of 83.65%, a precision of 88.32%, and a recall of 78.54%. Figure 3 shows the F1 scores of the different algorithms tested for smart home assistance for elderly people using the Cyborg system. The F1 score is the weighted harmonic mean of the precision and recall, with 0.0 being the worst and 1.0 being the best.

Figure 3 shows the F1 scores of the techniques used in the comparative analysis. Our proposed work produced the best result at 0.91, while the GNB algorithm produced the worst result at 0.61. The error rate of the smart home assistance system with the different algorithms is shown in

Table 3. Error rates.

Algorithm	MAE	MSE
GNB	0.295	0.088
ABC	0.321	0.379
KNN	0.273	0.218
KNN-ABC	0.154	0.056

.

From Table 3, it can be seen that our proposed KNN-ABC approach had the lowest error rate and produced better outcomes than the other algorithms. Figure 4 shows the correlation matrix for predicting intruders in the smart home environment.

In Figure 4, the diagonal values are not meaningful as they are self-correlated, i.e., with the variable itself. The values shown to the left and right of diagonal are considered mirror images of each other. The highly correlated variables are shown as darker boxes. Here, standing activities of intruders are highly correlated with one another. Therefore, detection of intruder with activity is predicted as sitting on the bed. Figure 5 shows the accuracy rate of the different tested techniques.

From Figure 5 shows the accuracy rates of the different techniques in the smart home environment system: for detection of intruders, GNB reached 88.12%; for power saving, Artificial Bee Colony algorithm (ABC) reached 90.12%; for determining the safest place to go in an emergency situation, the KNN technique reached an accuracy rate of 91.45%; finally, our proposed KNN-ABC approach reached 93.72%. Figure 6 shows the computation times for the various techniques tested in the smart home system.

Figure 6, shows that of the various techniques, our proposed KNN-ABC approach requires the lowest computation time.

Limitations and Future Work: This work considers a new hybrid approach, i.e., KNN-ABC, which combines the virtues of the KNN and ABC methods to improve on these methods as well as on the GNN method in terms of recall rate, precision, accuracy, F1 score, and computational complexity. However, the performance evaluation in this paper was restricted to smart home environment applications, and warrants further investigation to determine its improvement over other machine learning methods. Such an investigation and comparison that considers more advanced machine learning classifiers and data from other applications is recommended as a future extension of this work.

5. Conclusions

This paper proposes the implementation of a smart home environment control system through various sensor modules, control modules, and a human intervention robot named Cyborg; the system is used for control of the automatic ON/OFF state of various electronic appliances, detection of intruders in the smart home, and sending various alert notifications to the user. This system is intended to assist elderly people in an effective and efficient way with a fast response time. The accuracy rate of the various techniques was tested; GNB for the detection of intruders in the smart home environment system using the human-intervention robot named Cyborg reached 88.12%, the Artificial Bee Colony algorithm (ABC) for power saving in smart home electronic appliances reached 90.12%, the KNN technique for predicting safe locations in an emergency situation reached an accuracy rate of 91.45%, and our proposed KNN-ABC algorithm reached 93.72%. In the future, this work could be extended by providing the Cyborg robot system with more security features, including biometric concepts such as facial recognition and fingerprint identification. In summary, the proposed KNN-ABC algorithm shows better accuracy, precision, and recall rate than the GNB, KNN, and ABC algorithms. Better performance means more accurate prediction of activities and can lead to user satisfaction, a greater sense of security and safety, and improved quality of life.