Next Article in Journal
IoT-Enabled Sustainable and Cost-Efficient Returnable Transport Management Strategies in Multimodal Transport Systems
Next Article in Special Issue
Predicting Flood Hazards in the Vietnam Central Region: An Artificial Neural Network Approach
Previous Article in Journal
An Integrated Two-Dimension Linguistic Intuitionistic Fuzzy Decision-Making Approach for Unmanned Aerial Vehicle Supplier Selection
Previous Article in Special Issue
Understanding the Factors Influencing Consumers’ Intention toward Shifting to Solar Energy Technology for Residential Use in Saudi Arabia Using the Technology Acceptance Model
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:

Cloud-Based Fault Prediction for Real-Time Monitoring of Sensor Data in Hospital Environment Using Machine Learning

Chitkara University Institute of Engineering and Technology, Chitkara University, Punjab 140401, India
KIET Group of Institutions, Delhi NCR, Ghaziabad 201206, India
College of Computer Science and Information Systems, Najran University, Najran 61441, Saudi Arabia
Authors to whom correspondence should be addressed.
Sustainability 2022, 14(18), 11667;
Received: 14 August 2022 / Revised: 13 September 2022 / Accepted: 14 September 2022 / Published: 16 September 2022


The amount of data captured is expanding day by day which leads to the need for a monitoring system that helps in decision making. Current technologies such as cloud, machine learning (ML) and Internet of Things (IoT) provide a better solution for monitoring automation systems efficiently. In this paper, a prediction model that monitors real-time data of sensor nodes in a clinical environment using a machine learning algorithm is proposed. An IoT-based smart hospital environment has been developed that controls and monitors appliances over the Internet using different sensors such as current sensors, a temperature and humidity sensor, air quality sensor, ultrasonic sensor and flame sensor. The IoT-generated sensor data have three important characteristics, namely, real-time, structured and enormous amount. The main purpose of this research is to predict early faults in an IoT environment in order to ensure the integrity, accuracy, reliability and fidelity of IoT-enabled devices. The proposed fault prediction model was evaluated via decision tree, K-nearest neighbor, Gaussian naive Bayes and random forest techniques, but random forest showed the best accuracy over others on the provided dataset. The results proved that the ML techniques applied over IoT-based sensors are well efficient to monitor this hospital automation process, and random forest was considered the best with the highest accuracy of 94.25%. The proposed model could be helpful for the user to make a decision regarding the recommended solution and control unanticipated losses generated due to faults during the automation process.

1. Introduction

Automation plays a crucial part in the economic development of this golden era, which has an optimistic influence on the development of developed as well as developing nations [1,2]. The latest technologies are used by the automation industry to improve the economic competitiveness of the whole sector. Monitoring systems are an important part of the automation industry that help in enhancing production, lowering expense, early warning systems, predicting disease and much more [3,4,5,6]. Emerging technologies such as the Internet of Things, cloud and ML are used and incorporated with monitoring systems to enhance their performances. Many studies have proved the benefit of using IoT-based sensors in the monitoring process such as improving the working conditions, preventing errors, fault diagnosis, prediction of quality and helping managers make better decisions [7,8,9,10]. The amount of IoT-sensing devices is growing day by day; so, data generated from these are also growing exponentially.
Several earlier studies exhibited the benefits of the integration of IoT such as lowering the processing period in automation systems, feeding efficient solutions for the processing data yielded by IoT devices in smart cities and handling real-time data in a large amount of smart environments [11,12]. Data generated by various sensors are needed to be analyzed for effective decision making. Machine learning techniques are considered to be the latest technology having significant potential for the investigation of data, and they have been involved in different areas such as fault prediction, quality detection, defect classification and many more [13,14,15]. Machine learning techniques such as decision trees, naive Bayes and random forests are considered efficient in fault prediction by detecting abnormal events and preventing loss of productivity [16,17]. Cloud computing is the latest technology that has great possibilities for the business world. It is in great demand since 2000 and gives access to servers, databases, storage and other application services over the Internet [18]. In this era of digitalization, lifestyles are upgrading, but they are limited to the personal health of employees [19,20]. According to a report by the World Employment and Social Outlook, the total employment population equals 59%, out of which 33.33% of people work for more than forty-eight hours per week [21]. A huge amount of working people and increasing hours of working affect health in the short and long term. However, numerous investigations have proved that working for so long in the hospital has a severe effect on their health. A survey showed that 51.9% of registered health problem issues were due to long stays in clinical offices [22]. Moreover, in such cases, monitoring the health of employees efficiently using IoT has given a new path to maintaining optimal health in a clinical environment. The increasing amount of people resulted in the use of IoT, machine learning and cloud storage for transforming the concept of information perception, analysis and storage [23,24].
IoT systems generate a great amount of data in the automation industry, and as the data are complex, they become difficult to analyze. So, there is a need for using ML techniques to solve the diagnosis problems [25]. The decision tree (DT) is used to predict and classify discrete data. It predicts the value by learning from the season rules extracted from features of data [26]. The random forest (RF) machine learning method is applied to regression as well as classification problems. It is an ensemble learning method that creates many decision tree models [27]. The K-nearest neighbor (KNN) machine learning algorithm is easy to implement for both classification and regression problems. KNN is widely used in the IoT environment because of its computational efficiency and complexity [28]. The Gaussian Naive Bayes (GNB) algorithm is based on assumptions made by conditional independence and helps in analyzing data collected by the IoT system [29]. The major innovation and main goal of this work is to predict early faults in any IoT device in a hospital environment. The main shortcoming is data security on the cloud. Some research questions used within this research are:
RQ1: What are the main ML techniques used to predict the faults?
RQ2: What are the different performance metrics used in the proposed ML algorithm?
RQ3: What is the effect of the proposed work on the health of employees working in a clinical office?
There is no such deep study on the combination of IoT, cloud and machine learning techniques that monitor and predict faults in the automation system. Thus, a real-time fault prediction model using IoT-based sensors, cloud and machine learning algorithms is proposed for the hospital environment. The proposed model collects sensor data, whereas the cloud stores an enormous amount of fetched data. Finally, the proposed model consists of random forest classification to predict faults in the hospital automation process. The proposed model produces the following contributions:
  • A hospital automation system is designed and implemented in this paper that syncs the data to the cloud by analyzing the current values of the devices and sensors.
  • The real-time monitoring of data collected from different devices and sensors is completed for fault prediction.
  • A machine learning algorithm is designed that predicts faults in devices present in a hospital environment.
The rest of the paper is organized as follows: The extensive literature study is presented in Section 2. The methodology regarding the system design, proposed model and system implementation is illustrated in Section 3, whereas results and discussions are shown in Section 4. Section 5 contains the concluding remarks and the future scope.

2. Related Works

Emerging technologies such as sensors, machine learning and IoT are used to monitor and predict faults, improve production, reduce cost, provide early warning systems and help in taking better decisions for their management. This section examines the advantages and disadvantages of the work conducted by others related to the proposed model. Mariappan et al. [30] worked on security in the workplace via implementation of the RFID and GSM for a security-based control panel using a one-time password. Bhavana et al. [31] proposed a system that detects fault without interruption of humans by saving a lot of time. It was implemented using cloud, IoT and mobile application to monitor home appliances. Choudhary et al. [32] proposed the methodology for remotely controlled monitoring with minimum and secured configurations. The proposed model not just checks the sensors’ data information but also provides the solutions as per the situations, such as switching on the bulb in the low light. Sunehra and Tejaswi [33] proposed a model which used an ARM11 Raspberry Pi microcontroller unit that notifies the user about any interruption that occurs in the house in case the owner is away from the office/home. This automation model provides easiness, accessibility and reliability via control and real-time monitoring.
Alkar et al. [34] proposed an equipment/programming structure with minimal effort depending on IPAcBox that allows hospital/home frameworks to be handled via web association. In [35], the authors designed an architecture that described a model of smart hospital through IoT, cloud and machine learning technology for persons to increase their efficiency. This provides benefits such as security, easiness, energy consumption, costing and can be managed and monitored in the clinical environment easily. Kłosowski [36] provided a real-time scenario of deep learning (DL) methodologies for speech processing and modeling. Zhang et al. [37] deeply discussed the challenges of cloud computing and recognized the necessary research directions. In [38], the authors conducted an extensive literature review on recommendation techniques in the field of IoT. Some issues were also discussed which occurred during the implementation of the recommendation system.
Gladence et al. [39] understood the progress in automation and analyzed the capabilities of intelligent environments which are skilled in the detection of user-end actions and motions so that the corresponding assistance can be self-used for increasing the user’s convenience. Mandula et al. [40] and Milivojsa et al. [41] used the IoT in home automation with a microcontroller-based Arduino system and IOS/android app. In [42], the authors conducted an analysis of existing fault tolerance techniques applied in the automation industry. They also presented outliers’ prediction models used in the IoT, and sensor faults that occur in the automation system. Gobinath et al. [43] proposed an automation system which can switch on or off home devices such as the washing machine and AC using the IoT and AI. Katuk et al. [44] created a simple smart home automation system which includes smart devices such as mobile phones and firmware linked for data storage with a cloud server. Sarkar et al. [45] have designed a system that used an embedded framework via RFID to give a better and more efficient innovation used in the metro rail to encourage integrated ticketing, whereas Harsha et al. [46] designed a framework to provide a home/clinic automation system that connects with the end-user via different gestures through some parameters which are easily accessible.
Uppal et al. [47] used fall curve, a novel technique, to identify a fault in the sensor using its current value through the ACS712 current sensor. They also summarized different fault types and their causes. Balamurugan et al. [48] and Namasudra et al. [49] presented privacy and security problems that overcome malfunctioning issues in the cloud. In [50], the authors presented the role of AI in the software domain and discussed new technologies that save time and resources to enhance the quality. Abdul Ghaffar et al. and Malek et al. [51] proposed a combination of the IoT with big data to continuously monitor and process real-time data. In [52], the authors presented an IoT-based system focusing on the collection of data from sensors and its handling. Lo et al. [53] discussed ML tools used for fault diagnosis and detection in the industrial domain. In the era of industry 4.0, with the IoT, sensors develop a large volume and velocity of data, so, a cloud is essential for data storage. ML techniques are helpful to detect faults earlier rather than system failure. Mukhtar et al. [54] proposed a framework that uses sensors to check the health of a person carrying any disease, or not. Kocsis et al. [55] developed a worker-centric AI system in the SmartWork project that provides a good work environment for office workers. It also helps them to increase work efficiency and productivity of the employees through better management of the available resources. Rajput et al. [56] created a reference model for assisting diabetics in rural areas. It aids rural Indians in characterizing type 2 diabetes patients in their early stages. For the prediction, various ML classifiers were used, and their accuracy was compared to determine the best machine learning model. The SVM outperformed the other algorithms in terms of accuracy (96%). Rao et al. [57] concentrated on the management of energy in cubicles of a clinic/hospital. The authors proposed an IoT-based framework for vitality preservation in desk areas of the clinic/hospital.

3. Methodology

3.1. System Design

The developed system for real-time data monitoring and fault prediction provides support and identifies clinical/hospital devices as well as sends early notifications if a fault occurred [31]. This system uses current sensors, Arduino and a machine learning algorithm. The machine learning algorithm consists of the random forest classification model. As per Figure 1, there are multiple types of sensors attached to the control unit of the clinical/hospital design. These sensors are current, temperature and humidity, ultrasonic, air quality and fire sensors [58]. The data are captured by the control unit, which saves this data to the SD card, as well as transferred to the cloud server by the control unit. The cloud server processes the real-time data with a data processing unit and a machine learning algorithm for fault prediction and classification [59]. The machine learning algorithm has been trained over huge amounts of datasets. Therefore, it needs to process the data for fault prediction and classification which updates the data with the corresponding results. After fault prediction, the data are getting saved in the cloud which the end-user can access through the server via smartphone application. The fault prediction algorithm applies the random forest classification [60,61] to predict the faults in the incoming data during the process. In the end, all the historical sensor data related to current sensors, temperature and humidity, ultrasonic, air quality and fire are sent to the front end in real-time through a cloud server with fault prediction results [62].
The characteristics of data generated through device sensors are as follows: structured format, continuous generation and vast amount. Figure 2a depicts a sample of data yielded by device sensors in the JavaScript object notation layout and sent to the cloud server. The data generated by the sensor are being received at the back end for fault prediction and a machine algorithm is implemented for the same [63,64]. The model processes the data and calculates the fault result in the cloud server database. For fast read and write capacity, an embedded scheme-based technique is used. The final sensor data can be witnessed in Figure 2b, where sensor data of temperature and humidity, fire, door, switches, ac, coffee machine, printer, etc., fault prediction results are embedded as subdocuments.
In this paper, the fault prediction model is used to monitor the data fetched from different devices and sensors. The design of the model contains three rooms: one waiting lounge, one washroom and one pantry. Figure 3a shows the 2D layout of the clinical/hospital environment made with the help of the Planner 5D tool, whereas Figure 3b shows the 3D layout of the clinical/hospital environment made with the help of the Planner 5D tool.

3.2. Proposed Model for Fault Prediction

In this paper, a fault prediction machine learning algorithm has been proposed that checks whether the model is working fine or not. Figure 4 shows the process of gathering, preprocessing, classification and calculating results for the fetched data. In preprocessing, feature engineering and normalization techniques are applied to the fetched dataset. KNN, GNB, RF and DT techniques are applied in the classification of the dataset. The proposed model uses the random forest classification technique to identify and predict the classification category. For processing and calculating various prediction results, the dataset is collected from various sensors and devices. The collected dataset consists of around 30,000 instances which are classified into four cases: working fine, power supply off, need repair or need replacement during the process.
To evaluate the performance of different classification and prediction models, a dataset is collected from IoT-based sensors installed in different devices. In addition, the fetched dataset is labeled on the basis of different feasible combinations of faults that occur in the smart clinic/hospital. The machine learning technique learns and generates a robust model from the fetched dataset. Once the model is developed and installed in a smart clinic/hospital, the prediction of various devices from the real-time data of IoT sensors can be presented via mobile application. Once the dataset is received, fault data processing work is completed by removing inappropriate, irregular and missing data. Table 1 shows the dataset distribution for each class with respect to the mean, standard deviation (STD) and variance for every device.
Moreover, in order to analyze the identification of features, a confusion matrix is created. This distribution of the confusion matrix is shown in Table 2. The results of true negative (TN) and true positive (TP) are described as the count of accurately classified points. The results of false negative (FN) and false positive (FP) are described as the count of points inaccurately classified as “no” when it is actually a “yes” and classified as “yes” when it is actually a “no”.
Python is used to execute the classification models for the fetched dataset. Four classifications algorithms, namely, Gaussian naive Bayes, random forest, decision tree, K-nearest neighbor are applied to the dataset. Table 3 shows the performance metrics used for the proposed classification model to find the best one.
Figure 5 represents the flowchart of operations performed in the proposed system stepwise. Firstly, the data are sent from the devices to Arduino, and then Arduino sends the data of the devices to the server via a Wi-Fi unit. Cloud is used to manage a large amount of data of IoT devices and to control authorized access only. When the data are received at the server end, they will go through the machine learning algorithm for fault prediction. After prediction, it recommends solutions and saves the data to the cloud database. After that, the data from the cloud database are sent to the smartphone application that is accessible by the end-user.
Figure 6 represents the circuit diagram of the proposed clinic/hospital design in three stages, i.e., firstly, components used in the proposed model are shown; secondly, an intermediate stage; and in the third stage, a final circuit diagram with all connections are shown.
The prototype of the proposed model is made using a board and connected via glue as shown in Figure 7. Once the model is prepared, then it is populated with furniture, sensors and devices. There are a total of five rooms in the proposed model along with a sliding door and motion sensor on that door to detect the motion of a person. There are some switches on the model to induce the fault in devices.
Figure 8 shows the mobile application made for showing the results of fault prediction having different gauges. A gauge is a display widget that displays the values of different sensors present inside the clinic/hospital environment. The colors in Figure 8 define the status of the device. The green color symbolizes working fine, blue means needs to repair and red means a need to replace. In this, the solutions are also given according to the values received by the sensors and devices. A total of four cases are there, namely, working fine, can be repaired, needs replacement and main power supply off. The values send the output in one of the above cases which are represented by the mobile application, and the user is informed regarding the same.

4. Results

The visualization of data is conducted using the JavaScript framework to monitor real-time sensor data. The user can monitor the status of devices and will receive an early warning about the abnormality of their devices which is predicted in real-time through this proposed system. These IoT-based sensors send data to the cloud where an ML algorithm is applied to predict faults, and the results are directly forwarded to monitoring system to alert the user via mobile application. IoT-based sensors and devices are installed in the smart clinic/hospital and their data are transmitted to the server every three seconds. The proposed system contains three parts, namely, the IoT-based sensor, machine learning algorithm and prediction model. The evaluations of the performances of different algorithms are discussed in the next section. Figure 9a shows the instances of a sensor in all four categories having time in seconds on the x axis and current sensor value on the y axis. Figure 9b represents the image of the dataset formed to check the status of a device in each instance.
Figure 10 presents the flowchart of fault classification using a machine learning algorithm. Firstly, the dataset is taken and processed to analyze its parameters. An analysis of data is conducted using exploratory techniques to find out the different patterns of faults. Then, an ML classification algorithm is applied to classify the instances based on the patterns identified during the above techniques. Now after classification, malicious and normal devices are identified, and fault is predicted.
The different ML techniques used in this work are K-nearest neighbor, gaussian naive bayes, random forest and decision tree, and this answers RQ1. In machine learning, a confusion matrix, which is also known as an error matrix, is used to show the statistical classification of a problem. It is a matrix layout that helps in the visualization of the performance of the machine learning algorithm applied basically to supervised learning. Every row of the matrix depicts instances of the actual class, whereas every column of the matrix depicts instances of the predicted class. Through the confusion matrix, it becomes easy to see whether a system is confusing between two classes or not. The predicted class is present on the x axis and the true class is represented on the y axis of the confusion matrix. Figure 11 represents the confusion matrix of KNN, GNB, RF and DT.
Table 4 lists the recall, accuracy, precision and F1 score of faulty devices for four models: GNB, KNN, RF and DT. These four performance metrics answer RQ2. Accuracy tells the number of accurately predicted values out of all values. The precision metric calculates the probability of the accurately predicted faulty devices among the devices categorized as fault-prone. A smaller amount of accurately predicting faulty devices or a larger amount of inaccurately labeled fault-free devices will result in a lower precision rate. The recall represents the number of correct hits found and the F1 score is the weighted average of recall and precision. Using GNB and KNN, an accuracy of 92.5% and 92.75% are achieved, respectively, which exhibit an improvement in accuracy after applying DT. The random forest algorithm produces an accuracy of 94.25%, recall of 94.25%, precision of 91.95% and the F1 score of 91.99%.

5. Discussion

Four classification models named KNN, GNB, RF and DT are used to evaluate the dataset using different metrics. To handle large volumes of data and complex training of the systems, deep learning can be used in the future. Figure 12 illustrates the metric values of different machine learning models. RF has the highest accuracy of 94.25% and recall as compared to others. The precision and F1 score of RF is a little bit less than DT, but due to a small difference between values, RF is considered the best based on the accuracy metric.
The proposed work had a positive effect on the health of employees working in a clinic/hospital. The early prediction of faults in real-time devices lessens the hassle caused due to device damage, cost and time to repair. It ensures indoor comfort and user satisfaction, and this answers RQ3.
The analysis of the receiver operating characteristic (ROC) curve is conducted to estimate the performance of a classifier model. The ROC curve presents a visual tradeoff between the classifier’s capability to accurately detect the faulty device and the non-faulty device. The area under ROC (AUC) curve acts as a numeric metric for the evaluation of the performance of various classification models on the basis of the same dataset and it is directly related to the ROC curve. The AUC curve helps to differentiate the performances of different ML classification techniques on the same dataset. The curve that is closer to the top-left corner gives the better performance. Figure 13 depicts ROC curves for different classifiers, namely, KNN, GNB, RF and DT.
The classifier that produces the ROC curve closer to the top-left corner performs better as compared to the curve along with the diagonal. The closer the curve is to the 45-degree diagonal, the less accurate the classifier. The plots in the graph shown in Figure 13d are closest to the top-left corner which shows their best accuracy in detecting faulty and non-faulty devices. The ROC curve for classifier DT is the closest to the diagonal which shows its least accuracy in detecting faulty and non-faulty devices, whereas the ROC curve for classifier KNN and ROC curve for classifier GNB perform more accurately than DT but less accurately than RF. Thus, the ROC curve for classifier RF shows its maximum accuracy as compared to other classifiers. The range of current (in ampere) for all devices in our model is short. For example, if the average current value is 9 A, the permissible range lies between 8 A and 10 A, which means, the fluctuation occurring in any device is within the range +1 to −1. Therefore, in our dataset, different sensors are within a pre-defined range, which causes minimal statistical fluctuations. Thus, the ROC curves are plotted as straight lines.
A precision–recall (PR) curve offers a substitute method for visual comparison of different classifiers. A PR curve can expose the dissimilarity between different algorithms which is not possible from a ROC curve. In a PR curve, the y axis is precision and the x axis represents recall.
Figure 14 shows the PR curves for KNN, GNB, RF and DT, respectively. The random forest classification algorithm achieves better results than others, as a large area under a PR curve proves high precision and high recall. More precision means a low false positive rate and a high recall means a low false negative rate. All devices in our model have a limited current range (in amperes). For example, if the average current value is 3 A, the permissible range is 2 A to 4 A, implying that fluctuation in any device is between +1 and −1. As a result, different sensors in our dataset are within a pre-defined range, resulting in minimal statistical fluctuations. The key in calculating precision and recall is that it does not use true negatives. It is only concerned with predicting the minority class correctly. In our dataset, there is a significant class imbalance as the non-faulty cases (off and fine classes) are found to be the most, and faulty cases (repair and replace classes) are minimal. Furthermore, in PR curves, the accuracy is said to be the most if the precision and recall values tend to ‘1’. In the graph presented in Figure 14, the majority of cases (non-faulty cases), i.e., the ‘OFF’ and ‘FINE’ classes are represented in blue and yellow plots, respectively. The plots of both cases have reached the value ‘1’. The major challenge of the proposed work is privacy and the security of training data present on the cloud. However, computing a false positive is a major threat to the validity of the results.

6. Conclusions

In this paper, the authors developed a real-time fault prediction model that uses IoT-based sensors, cloud and machine learning algorithms. The proposed model predicts faults and prevents unexpected casualties caused by faults. Through this study, the authors showed that the integration of IoT-based sensors with the cloud is efficient for investigating and processing a large amount of real-time data. In addition, the performance of the model was analyzed with various metrics such as accuracy, recall, F1 score and precision. In all experiment cases, IoT-based sensors gave an efficient recommendation as they gathered and communicated data successfully in less duration and cost. The real-time monitoring of data was achieved through remote access with minimum cost and secure authentication. The techniques used in this paper have achieved the target of real-time clinical/hospital device monitoring and notifications through the Wi-Fi module. This model not only shows the sensor data-related information, but also shows the information related to the status of each device at every timestamp whether the devices are working fine, need to be repaired, need to be replaced or are in the off mode.
Fault prediction is important as it can easily predict whether a device is working fine or not. All three research questions have been answered in the manuscript. A total of four prediction models were chosen, out of which random forest performed best with the highest accuracy of 94.25%, followed by K-nearest neighbor having an accuracy of 92.75%. Decision tree proved to be the least preferable for fault prediction in a clinic/hospital environment in the fetched dataset. The authors proposed a fault prediction model that uses a random forest classification technique and cloud. The results proved that the developed model is efficient with the highest accuracy as compared to another model. The approach and steps used for evaluation in this paper are apt for a better understanding; therefore, for future research, they help in enhancing the statistical validity. In the future, security is a major issue when there are a large number of IoT devices; therefore, the security of IoT devices should also be considered. A key future challenge is regarding the inclusion of a greater number of IoT devices connected to each other in a clinical/hospital environment. So, in this case, deep learning can be applied depending on the size of the dataset.

Author Contributions

Conceptualization, M.U., D.G., S.J. and A.S. (Adel Sulaiman); methodology, A.R., K.R., M.A.E. and A.S. (Asadullah Shaikh); software, M.U., D.G. and A.S. (Adel Sulaiman); validation, K.R., A.R. and A.S. (Asadullah Shaikh); formal analysis, S.J. and K.R.; investigation, M.U. and A.S. (Asadullah Shaikh); resources, A.S. (Adel Sulaiman) and A.R.; data curation, D.G., M.A.E. and S.J.; writing—original draft preparation, M.U., D.G., S.J. and A.S. (Adel Sulaiman); writing—review and editing, A.R., K.R., M.A.E. and A.S. (Asadullah Shaikh); visualization, A.S. (Adel Sulaiman); supervision, A.S. (Asadullah Shaikh); project administration, M.U. and S.J.; funding acquisition, A.S. (Adel Sulaiman). All authors have read and agreed to the published version of the manuscript.


The authors would like to thank the Deputy for Study and Innovation, Ministry of Education, Kingdom of Saudi Arabia, for funding this research through a grant (NU/IFC/INT/01/008) from the Najran University Institutional Funding Committee.

Data Availability Statement

All the data is available within the article.

Conflicts of Interest

The authors declare no conflict of interest.


  1. Prettner, K. A note on the implications of automation for economic growth and the labor share. Macroecon. Dyn. 2019, 23, 1294–1301. [Google Scholar] [CrossRef]
  2. Acemoglu, D.; Restrepo, P. Secular stagnation? The effect of aging on economic growth in the age of automation. Am. Econ. Rev. 2017, 107, 174–179. [Google Scholar] [CrossRef]
  3. Mora, H.; Gil, D.; Terol, R.M.; Azorín, J.; Szymanski, J. An IoT-Based Computational Framework for Healthcare Monitoring in Mobile Environments. Sensors 2017, 17, 2302. [Google Scholar] [CrossRef]
  4. Zhang, X.; Zhang, J.; Li, L.; Zhang, Y.; Yang, G. Monitoring Citrus Soil Moisture and Nutrients Using an IoT Based System. Sensors 2017, 17, 447. [Google Scholar] [CrossRef] [PubMed]
  5. Manes, G.; Collodi, G.; Gelpi, L.; Fusco, R.; Ricci, G.; Manes, A.; Passafiume, M. Realtime Gas Emission Monitoring at Hazardous Sites Using a Distributed Point-Source Sensing Infrastructure. Sensors 2016, 16, 121. [Google Scholar] [CrossRef] [PubMed]
  6. Cheung, W.-F.; Lin, T.-H.; Lin, Y.-C. A Real-Time Construction Safety Monitoring System for Hazardous Gas Integrating Wireless Sensor Network and Building Information Modeling Technologies. Sensors 2018, 18, 436. [Google Scholar] [CrossRef] [PubMed]
  7. Moon, Y.S.; Choi, H.R.; Kim, J.J.; Kim, D.W.; Cho, J.H.; Kim, J.W.; Jeong, J.W. Development of IoT-Based Sensor Tag for Smart Factory. Int. Res. J. Electron. Comput. Eng. 2017, 3, 28–31. [Google Scholar]
  8. Li, J.; Xie, J.; Yang, Z.; Li, J. Fault Diagnosis Method for a Mine Hoist in the Internet of Things Environment. Sensors 2018, 18, 1920. [Google Scholar] [CrossRef]
  9. Lee, J.; Noh, S.D.; Kim, H.-J.; Kang, Y.-S. Implementation of Cyber-Physical Production Systems for Quality Prediction and Operation Control in Metal Casting. Sensors 2018, 18, 1428. [Google Scholar] [CrossRef]
  10. Mittal, S.; Bansal, A.; Gupta, D.; Juneja, S.; Turabieh, H.; Elarabawy, M.M.; Bitsue, Z.K. Using Identity-Based Cryptography as a Foundation for an Effective and Secure Cloud Model for E-Health. Comput. Intell. Neurosci. 2022, 2022, 7016554. [Google Scholar] [CrossRef]
  11. Lohokare, J.; Dani, R.; Rajurkar, A.; Apte, A. An IoT ecosystem for the implementation of scalable wireless home automation systems at smart city level. In Proceedings of the IEEE Region 10 Conference, Penang, Malaysia, 5–8 November 2017; pp. 1503–1508. [Google Scholar]
  12. Villari, M.; Celesti, A.; Fazio, M.; Puliafito, A. AllJoyn Lambda: An architecture for the management of smart environments in IoT. In Proceedings of the International Conference on Smart Computing Workshops, Hong Kong, China, 5 November 2014; pp. 9–14. [Google Scholar]
  13. Correa-Baena, J.P.; Hippalgaonkar, K.; van Duren, J.; Jaffer, S.; Chandrasekhar, V.R.; Stevanovic, V.; Wadia, C.; Guha, S.; Buonassisi, T. Accelerating materials development via automation, machine learning, and high-performance computing. Joule 2018, 2, 1410–1420. [Google Scholar] [CrossRef][Green Version]
  14. Chen, Y.J.; Fan, C.Y.; Chang, K.H. Manufacturing intelligence for reducing false alarm of defect classification by integrating similarity matching approach in CMOS image sensor manufacturing. Comput. Ind. Eng. 2016, 99, 465–473. [Google Scholar] [CrossRef]
  15. Ravikumar, S.; Ramachandran, K.I.; Sugumaran, V. Machine learning approach for automated visual inspection of machine components. Expert Syst. 2011, 38, 3260–3266. [Google Scholar] [CrossRef]
  16. Patel, R.K.; Giri, V.K. Feature selection and classification of mechanical fault of an induction motor using random forest classifier. Perspect. Sci. 2016, 8, 334–337. [Google Scholar] [CrossRef]
  17. Cerrada, M.; Zurita, G.; Cabrera, D.; Sánchez, R.V.; Artés, M.; Li, C. Fault diagnosis in spur gears based on genetic algorithm and random forest. Mech. Syst. Signal Process. 2016, 70, 87–103. [Google Scholar] [CrossRef]
  18. Wu, J.; Ping, L.; Ge, X.; Wang, Y.; Fu, J. Cloud storage as the infrastructure of cloud computing. In Proceedings of the International Conference on Intelligent Computing and Cognitive Informatics, Kuala Lumpur, Malaysia, 22–23 June 2010; pp. 380–383. [Google Scholar]
  19. He, C.; Fan, X.; Li, Y. Toward ubiquitous healthcare services with a novel efficient cloud platform. Trans. Biomed. Eng. 2013, 60, 230–234. [Google Scholar] [CrossRef] [PubMed]
  20. Hossain, M.S.; Muhammad, G. Cloud-assisted industrial internet of things (iiot)–enabled framework for health monitoring. Comput. Netw. 2016, 101, 192–202. [Google Scholar] [CrossRef]
  21. Kour, K.; Gupta, D.; Gupta, K.; Juneja, S.; Kaur, M.; Alharbi, A.H.; Lee, H.N. Controlling Agronomic Variables of Saffron Crop Using IoT for Sustainable Agriculture. Sustainability 2022, 14, 5607. [Google Scholar] [CrossRef]
  22. Yu, W.; Lao, X.Q.; Pang, S.; Zhou, J.; Zhou, A.; Zou, J.; Mei, L.; Yu, I.T. A survey of occupational health hazards among 7610 female workers in Chinas electronics industry. Arch. Environ. Occup. Health 2013, 68, 190–205. [Google Scholar] [CrossRef]
  23. Sharma, S.; Gupta, K.; Gupta, D.; Juneja, S.; Turabieh, H.; Sharma, S. SWOT: A Hybrid Hardware-Based Approach for Robust Fault-Tolerant Framework in a Smart Day Care. Secur. Commun. Netw. 2022, 2022, 2939469. [Google Scholar] [CrossRef]
  24. Xu, B.; Xu, L.; Cai, H.; Jiang, L.; Luo, Y.; Gu, Y. The design of an m-Health monitoring system based on a cloud computing platform. Enterp. Inf. Syst. 2017, 11, 17–36. [Google Scholar] [CrossRef]
  25. Ademujimi, T.T.; Brundage, M.P.; Prabhu, V.V. A review of current machine learning techniques used in manufacturing diagnosis. In IFIP International Conference on Advances in Production Management Systems; Springer: Cham, Switzerland, 2017; pp. 407–415. [Google Scholar]
  26. Jahromi, A.T.; Er, M.J.; Li, X.; Lim, B.S. Sequential fuzzy clustering based dynamic fuzzy neural network for fault diagnosis and prognosis. Neurocomputing 2016, 196, 31–41. [Google Scholar] [CrossRef]
  27. Singh, P.; Kaur, R.; Rashid, J.; Juneja, S.; Dhiman, G.; Kim, J.; Ouaissa, M. A Fog-Cluster Based Load-Balancing Technique. Sustainability 2022, 14, 7961. [Google Scholar] [CrossRef]
  28. Alam, F.; Mehmood, R.; Katib, I.; Albeshri, A. Analysis of eight data mining algorithms for smarter Internet of Things (IoT). Procedia Comput. Sci. 2016, 98, 437–442. [Google Scholar] [CrossRef]
  29. Zhang, Z.; Zhang, S. Application of Internet of Things and Naive Bayes in Public Health Environmental Management of Government Institutions in China. J. Healthc. Eng. 2021, 2021, 9171756. [Google Scholar] [CrossRef] [PubMed]
  30. Mariappan, D.B.; Parihar, H.; Gautham, M.K.; Verma, D.C. Smart Clinical/Hospital Area Monitoring & Control Based on IOT. In Proceedings of the 2020 2nd International Conference on Advances in Computing, Communication Control and Networking (ICACCCN), Greater Noida, India, 18–19 December 2020; pp. 450–453. [Google Scholar]
  31. Bhavana, K.; Nekkanti, V.; Jayapandian, N. Internet of things enabled device fault prediction system using machine learning. In International Conference on Inventive Computation Technologies; Springer: Cham, Switzerland, 2019; pp. 920–927. [Google Scholar]
  32. Choudhary, V.; Parab, A.; Bhapkar, S.; Jha, N.; Kulkarni, M. Desgin and implementation of wi-fi based smart home system. Int. J. Eng. Comput. Sci. 2016, 5, 400022. [Google Scholar] [CrossRef]
  33. Sunehra, D.; Tejaswi, V. Implementation of speech based home automation system using Bluetooth and GSM. In Proceedings of the 2016 International Conference on Signal Processing, Communication, Power and Embedded System (SCOPES), Paralakhemundi, India, 3–5 October 2016; pp. 807–813. [Google Scholar]
  34. Alkar, A.Z.; Roach, J.; Baysal, D. IP based home automation system. IEEE Trans. Consum. Electron. 2010, 56, 2201–2207. [Google Scholar] [CrossRef]
  35. Uppal, M.; Gupta, D.; Juneja, S.; Dhiman, G.; Kautish, S. Cloud-Based Fault Prediction Using IoT in Clinical/Hospital Automation for Improvisation of Health of Employees. J. Healthc. Eng. 2021. [Google Scholar] [CrossRef]
  36. Kłosowski, P. Deep learning for natural language processing and language modelling. In Proceedings of the 2018 Signal Processing: Algorithms, Architectures, Arrangements, and Applications (SPA), Poznan, Poland, 19–21 September 2018; pp. 223–228. [Google Scholar]
  37. Zhang, Q.; Cheng, L.; Boutaba, R. Cloud computing: State-of-the-art and research challenges. Journal of internet services and applications. J. Internet Serv. Appl. 2010, 1, 7–18. [Google Scholar] [CrossRef]
  38. Kanwal, S.; Rashid, J.; Kim, J.; Juneja, S.; Dhiman, G.; Hussain, A. Mitigating the coexistence technique in wireless body area networks by using superframe interleaving. IETE J. Res. 2022, 1–15. [Google Scholar] [CrossRef]
  39. Gladence, L.M.; Karthi, M.; Anu, V.M. A statistical comparison of logistic regression and different Bayes classification methods for machine learning. ARPN J. Eng. Appl. Sci. 2015, 10, 5947–5953. [Google Scholar]
  40. Mandula, K.; Parupalli, R.; Murty, C.A.; Magesh, E.; Lunagariya, R. Mobile based home automation using Internet of Things (IoT). In Proceedings of the 2015 International Conference on Control, Instrumentation, Communication and Computational Technologies (ICCICCT), Kumaracoil, India, 18–19 December 2015; pp. 340–343. [Google Scholar]
  41. Milivojša, S.; Ivanović, S.; Erić, T.; Antić, M.; Smiljković, N. Implementation of voice control interface for smart home automation system. In Proceedings of the 2017 IEEE 7th International Conference on Consumer Electronics-Berlin (ICCE-Berlin), Berlin, Germany, 3–6 September 2017; pp. 263–264. [Google Scholar]
  42. Sharma, S.; Gupta, S.; Gupta, D.; Juneja, S.; Gupta, P.; Dhiman, G.; Kautish, S. Deep Learning Model for the Automatic Classification of White Blood Cells. Comput. Intell. Neurosci. 2022, 2022, 7384131. [Google Scholar] [CrossRef] [PubMed]
  43. Gobinath, V.; Gladence, L.M.; Haran, S.H.; Sathiyavathi, R.; Anu, V.M. Home and clinical/Hospital automation system-survey. In Proceedings of the 2017 Third International Conference on Science Technology Engineering & Management (ICONSTEM), Chennai, India, 23–24 March 2017; pp. 435–440. [Google Scholar]
  44. Katuk, N.; Ku-Mahamud, K.R.; Zakaria, N.H.; Maarof, M.A. Implementation and recent progress in cloud-based smart home automation systems. In Proceedings of the 2018 IEEE Symposium on Computer Applications & Industrial Electronics (ISCAIE), Penang, Malaysia, 28–29 April 2018; pp. 71–77. [Google Scholar]
  45. Sarkar, D.; Patel, H.; Dave, B. Development of integrated cloud-based Internet of Things (IoT) platform for asset management of elevated metro rail projects. Int. J. Constr. Manag. 2020, 22, 1993–2002. [Google Scholar] [CrossRef]
  46. Harsha, S.S.; Reddy, S.C.; Mary, S.P. Enhanced home automation system using Internet of Things. In Proceedings of the 2017 International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC), Palladam, India, 10–11 February 2017; pp. 89–93. [Google Scholar]
  47. Uppal, M.; Gupta, D.; Anand, D.; Alharithi, F.S.; Almotiri, J.; Mansilla, A.; Singh, D.; Goyal, N. Fault Pattern Diagnosis and Classification in Sensor Nodes Using Fall Curve. Comput. Mater. Contin. 2022, 72, 1799–1814. [Google Scholar] [CrossRef]
  48. Balamurugan, B.; Venkata Krishna, P.; Kumar, N.S.; Rajyalakshmi, G.V. An efficient framework for health system based on hybrid cloud with ABE-outsourced decryption. In Artificial Intelligence and Evolutionary Algorithms in Engineering Systems; Springer: Berlin/Heidelberg, Germany, 2015; pp. 41–49. [Google Scholar]
  49. Namasudra, S.; Roy, P. Secure and efficient data access control in cloud computing environment: A survey. Multiagent Grid Syst. 2016, 12, 69–90. [Google Scholar] [CrossRef]
  50. Kour, K.; Gupta, D.; Gupta, K.; Dhiman, G.; Juneja, S.; Viriyasitavat, W.; Islam, M.A. Smart-Hydroponic-Based Framework for Saffron Cultivation: A Precision Smart Agriculture Perspective. Sustainability 2022, 14, 1120. [Google Scholar] [CrossRef]
  51. Malek, Y.N.; Kharbouch, A.; El Khoukhi, H.; Bakhouya, M.; De Florio, V.; El Ouadghiri, D.; Latré, S.; Blondia, C. On the use of IoT and big data technologies for real-time monitoring and data processing. Procedia Comput. Sci. 2017, 113, 429–434. [Google Scholar] [CrossRef]
  52. AbdulGhaffar, A.; Mostafa, S.M.; Alsaleh, A.; Sheltami, T.; Shakshuki, E.M. Internet of things based multiple disease monitoring and health improvement system. J. Ambient. Intell. Humaniz. Comput. 2020, 11, 1021–1029. [Google Scholar] [CrossRef]
  53. Lo, N.G.; Flaus, J.M.; Adrot, O. Review of machine learning approaches in fault diagnosis applied to IoT systems. In Proceedings of the 2019 International Conference on Control, Automation and Diagnosis (ICCAD), Grenoble, France, 2–4 July 2019; pp. 1–6. [Google Scholar]
  54. Mukhtar, H.; Rubaiee, S.; Krichen, M.; Alroobaea, R. An IoT framework for screening of COVID-19 using real-time data from wearable sensors. Int. J. Environ. Res. Public Health 2021, 18, 4022. [Google Scholar] [CrossRef]
  55. Kocsis, O.; Moustakas, K.; Fakotakis, N.; Vassiliou, C.; Toska, A.; Vanderheiden, G.C.; Stergiou, A.; Amaxilatis, D.; Pardal, A.; Quintas, J.; et al. Smartwork: Designing a smart age-friendly living and working environment for clinical/Hospital workers. In Proceedings of the 12th ACM International Conference on Pervasive Technologies Related to Assistive Environments, Rhodes Greece, 5–7 June 2019; pp. 435–441. [Google Scholar]
  56. Rajput, D.S.; Basha, S.M.; Xin, Q.; Gadekallu, T.R.; Kaluri, R.; Lakshmanna, K.; Maddikunta, P.K.R. Providing diagnosis on diabetes using cloud computing environment to the people living in rural areas of India. J. Ambient. Intell. Humaniz. Comput. 2022, 13, 2829–2840. [Google Scholar] [CrossRef]
  57. Rao, M.V.; Krishna, T.R.; Ganduri, R.; Roohi, A. An effective energy management system for smart clinical/Hospital cubicles using IoT. J. Adv. Res. Dyn. Control. Syst. 2018, 10, 338–347. [Google Scholar]
  58. Sharma, S.; Gupta, S.; Gupta, D.; Juneja, S.; Singal, G.; Dhiman, G.; Kautish, S. Recognition of Gurmukhi Handwritten City Names Using Deep Learning and Cloud Computing. Sci. Program. 2022, 2022, 5945117. [Google Scholar] [CrossRef]
  59. Gupta, N.; Gupta, K.; Gupta, D.; Juneja, S.; Turabieh, H.; Dhiman, G.; Kautish, S.; Viriyasitavat, W. Enhanced virtualization-based dynamic bin-packing optimized energy management solution for heterogeneous clouds. Math. Probl. Eng. 2022, 2022, 8734198. [Google Scholar] [CrossRef]
  60. Monga, C.; Gupta, D.; Prasad, D.; Juneja, S.; Muhammad, G.; Ali, Z. Sustainable Network by Enhancing Attribute-Based Selection Mechanism Using Lagrange Interpolation. Sustainability 2022, 14, 6082. [Google Scholar] [CrossRef]
  61. Ray, P.; Kaluri, R.; Reddy, T.; Lakshmanna, K. Contemporary Developments and Technologies in Deep Learning–Based IoT. In Deep Learning for Internet of Things Infrastructure; CRC Press: Boca Raton, FL, USA, 2021; pp. 61–82. [Google Scholar]
  62. Dev, K.; Maddikunta, P.K.R.; Gadekallu, T.R.; Bhattacharya, S.; Hegde, P.; Singh, S. Energy optimization for green communication in IoT using harris hawks optimization. IEEE Trans. Green Commun. Netw. 2022, 6, 685–694. [Google Scholar] [CrossRef]
  63. Iwendi, C.; Maddikunta, P.K.R.; Gadekallu, T.R.; Lakshmanna, K.; Bashir, A.K.; Piran, M.J. A metaheuristic optimization approach for energy efficiency in the IoT networks. Softw. Pract. Exp. 2021, 51, 2558–2571. [Google Scholar] [CrossRef]
  64. Dhiman, G.; Rashid, J.; Kim, J.; Juneja, S.; Viriyasitavat, W.; Gulati, K. Privacy for Healthcare Data Using the Byzantine Consensus Method. IETE J. Res. 2022, 1–12. [Google Scholar] [CrossRef]
Figure 1. Proposed architecture for fault prediction in clinical/hospital environment.
Figure 1. Proposed architecture for fault prediction in clinical/hospital environment.
Sustainability 14 11667 g001
Figure 2. Sample of generated sensor data (a) demonstrated in JSON format (b) and saved in NoSQL DB.
Figure 2. Sample of generated sensor data (a) demonstrated in JSON format (b) and saved in NoSQL DB.
Sustainability 14 11667 g002
Figure 3. Design using Planner 5D tool: (a) in 2D layout format; (b) in 3D layout format.
Figure 3. Design using Planner 5D tool: (a) in 2D layout format; (b) in 3D layout format.
Sustainability 14 11667 g003
Figure 4. Block diagram of proposed clinical/hospital environment.
Figure 4. Block diagram of proposed clinical/hospital environment.
Sustainability 14 11667 g004
Figure 5. Flowchart of operations performed in proposed system.
Figure 5. Flowchart of operations performed in proposed system.
Sustainability 14 11667 g005
Figure 6. Circuit diagram of proposed clinical/hospital design. (a) Components used in proposed model. (b) Intermediate stage. (c) Final circuit diagram with all connections.
Figure 6. Circuit diagram of proposed clinical/hospital design. (a) Components used in proposed model. (b) Intermediate stage. (c) Final circuit diagram with all connections.
Sustainability 14 11667 g006
Figure 7. Implementation of proposed smart clinical/hospital prototype.
Figure 7. Implementation of proposed smart clinical/hospital prototype.
Sustainability 14 11667 g007
Figure 8. Solution recommendations using mobile application.
Figure 8. Solution recommendations using mobile application.
Sustainability 14 11667 g008
Figure 9. The real-time data monitoring of data fetched from a sensor for all four categories. (a) Instances of a sensor in all four categories. (b) Screenshot of dataset to check the status of a device in each instance.
Figure 9. The real-time data monitoring of data fetched from a sensor for all four categories. (a) Instances of a sensor in all four categories. (b) Screenshot of dataset to check the status of a device in each instance.
Sustainability 14 11667 g009
Figure 10. Flowchart of different operations applied in machine learning algorithm.
Figure 10. Flowchart of different operations applied in machine learning algorithm.
Sustainability 14 11667 g010
Figure 11. Confusion matrix for KNN, GNB, RF and DT, respectively. (a) DT trained model confusion matrix. (b) GNB trained model confusion matrix. (c) KNN trained model confusion matrix. (d) RF trained model confusion matrix.
Figure 11. Confusion matrix for KNN, GNB, RF and DT, respectively. (a) DT trained model confusion matrix. (b) GNB trained model confusion matrix. (c) KNN trained model confusion matrix. (d) RF trained model confusion matrix.
Sustainability 14 11667 g011
Figure 12. Performance comparison of DT, GNB, KNN and RF.
Figure 12. Performance comparison of DT, GNB, KNN and RF.
Sustainability 14 11667 g012
Figure 13. Receiver operating characteristic curves for KNN, GNB, RF and DT, respectively. (a) ROC curve for classifier KNN. (b) ROC curve for classifier DT. (c) ROC curve for classifier GNB. (d) ROC curve for classifier RF.
Figure 13. Receiver operating characteristic curves for KNN, GNB, RF and DT, respectively. (a) ROC curve for classifier KNN. (b) ROC curve for classifier DT. (c) ROC curve for classifier GNB. (d) ROC curve for classifier RF.
Sustainability 14 11667 g013
Figure 14. Precision vs. recall curves for KNN, GNB, RF and DT, respectively. (a) PR curve for classifier KNN. (b) PR curve for classifier DT. (c) PR curve for classifier GNB. (d) PR curve for classifier RF.
Figure 14. Precision vs. recall curves for KNN, GNB, RF and DT, respectively. (a) PR curve for classifier KNN. (b) PR curve for classifier DT. (c) PR curve for classifier GNB. (d) PR curve for classifier RF.
Sustainability 14 11667 g014
Table 1. Distribution of fetched dataset.
Table 1. Distribution of fetched dataset.
Description ACCoffeeTVPrinterCCTVLight
POWER OFFMean0.00199790.00097830.00035490.00018340.00024680.0000876
Table 2. Confusion matrix.
Table 2. Confusion matrix.
Actual “Yes”Actual “No”
Classified as “Yes”TPFP
Classified as “No”FNTN
Table 3. Performance metrics applied to proposed classification model.
Table 3. Performance metrics applied to proposed classification model.
Performance MetricFormula
Specificity(TN)/(TN + FP)
Recall/SensitivityTP/(TP + FN)
False Positive Rate(FP)/(FP + TN)
True Positive RateTP/(TP + FN)
Accuracy(TP + TN)/(TP + TN + FP + FN)
PrecisionTP/(TP + FP)
F1-score(2 TP)/(2 TP + FP + FN)
Table 4. Performance metrics of different machine learning models for fault prediction.
Table 4. Performance metrics of different machine learning models for fault prediction.
ModelAccuracyPrecisionRecallF1 Score
Gaussian Naive Bayes92.508%91.806%92.508%91.936%
Random Forest94.257%91.951%94.257%91.990%
K-Nearest Neighbor92.753%91.994%92.753%92.210%
Decision Tree 91.963%92.061%91.963%92.010%
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Uppal, M.; Gupta, D.; Juneja, S.; Sulaiman, A.; Rajab, K.; Rajab, A.; Elmagzoub, M.A.; Shaikh, A. Cloud-Based Fault Prediction for Real-Time Monitoring of Sensor Data in Hospital Environment Using Machine Learning. Sustainability 2022, 14, 11667.

AMA Style

Uppal M, Gupta D, Juneja S, Sulaiman A, Rajab K, Rajab A, Elmagzoub MA, Shaikh A. Cloud-Based Fault Prediction for Real-Time Monitoring of Sensor Data in Hospital Environment Using Machine Learning. Sustainability. 2022; 14(18):11667.

Chicago/Turabian Style

Uppal, Mudita, Deepali Gupta, Sapna Juneja, Adel Sulaiman, Khairan Rajab, Adel Rajab, M. A. Elmagzoub, and Asadullah Shaikh. 2022. "Cloud-Based Fault Prediction for Real-Time Monitoring of Sensor Data in Hospital Environment Using Machine Learning" Sustainability 14, no. 18: 11667.

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop