PM2.5 pollution is dangerous air pollution that threatens public health. WSN-based PM2.5 pollutant prediction system is a solution to increase public awareness of its long-term effects. However, outliers and sparse data are common in datasets downloaded from a large-scale WSN causing the accuracy of a prediction system to decrease. A prediction model based on convolutional long-short term memory (ConvLSTM) is proposed in this study. This model is built by combining the convolution layer and the long short-term memory layer. The datasets are mined from a network of sensors spread across Taiwan. Using the last 24-hour records, the next hour PM2.5 concentration is predicted. The root mean square error (RMSE) was calculated to evaluate the prediction accuracy. It can be seen that the inference model that uses convLSTM produces better performance than the one using ARIMA regression or conventional neural network.

Agriculture is one sector that has a major role in the national economy, where agriculture has a contribution of 13.53 percent of GPA in Indonesia. The most widely produced agricultural products are types of grains such as corn, rice, and beans. However, the price of these commodities is often controlled by brokers, several aspects that become the benchmark for brokers to price these commodities are the quality of the seeds. The quality of rice or corn seeds is considered good if it meets a very small level of moisture content. Therefore, sometimes farmers dry their harvests for a long time without knowing the level of water content contained in the nature of their harvests so that it will hamper the distribution of crops. The solution to the problems experienced by the farmers is a device that can detect the moisture content in harvested seeds at low prices and an easy-to-use process, and is supported by an AI system that can assist farmers in making decisions about crop quality. This device works by reading the level of moisture in the grain using a humidity sensor that is plugged into the harvest container such as sacks and others. If it is felt that the water level is low, the device will provide suggestions to the user to support decision making through the OLED screen installed on the device. The features make it easier for users to make decisions whether the harvest is worth selling or not.

One way to increase immunity and maintain immunity can be done by consuming herbal plants. This herbal medicine is empirically believed to be useful as a cultural treasure from generation to generation. All parts of the plant can be used as medicine, one of which is the leaves. However, most people do not know the herbal leaves. This herbal leaf can actually be recognized from the characteristics of its shape. This study aims to identify types of herbal leaves using first-order feature extraction and the Multiclass Support Vector Machine (Multiclass SVM) algorithm. First-order feature extraction is able to extract features using the parameters of mean, skewness, variance, kurtosis, and entropy. Meanwhile, Multiclass SVM identifies by obtaining the optimal line in separating the data set of two classes of two-dimensional space points in order to find the maximum hyperplane in separating the data points into classes so that they can be grouped. From the test results, the identification accuracy is an average of 76%. This shows that the algorithm has been able to identify, but needs improvement.

Some people consider traditional food and cakes to be out of date. Many of the traditional snacks were abandoned by the community and began to switch to more modern foods so that people sometimes do not recognize the traditional cakes in circulation. Due to the lack of dissemination of the introduction of this traditional snack, a system is needed to recognize traditional Indonesian snacks. As technology develops, image recognition using the Convolutional Neural Network (CNN) method as a classification method using the pre-trained MobilenetV2 model as the basic model can be used. From a total dataset of 1545 images of traditional cakes consisting of 8 categories, they are divided into 80% train data and 20% test data. After going through the training and testing process, the accuracy results are 98.9% for train data and 90.5% for test data. Model testing performed on the new test data resulted in an accuracy of 92.5% where the model managed to classify 74 images from 80 images of traditional cakes according to their categories which were presented in the form of a confusion matrix. Several experiments were also carried out to find the parameters that produce the model with the best accuracy, namely the effect of the number of epochs, the effect of the dataset distribution scenario, and the effect of the size of the learning rate.

Along with technology development, the classification system for leukemia is built using a computer base to assist health experts as a second opinion to make efficient and accurate diagnoses. This research was conducted by designing an image processing system for two types of images, normal and acute, by applying the Gray Level Co-occurrence Matrix (GLCM) feature extraction method and two classification methods: K-Nearest Neighbors (KNN) and Support Vector Machine (SVM). The design of this system aims to determine how effective the method is to classify leukemia images in terms of accuracy. The KNN method could classify with an accuracy rate of 96.7% at quantization level 8 with pixel distances of 50, 60, and 70. Conversely, the SVM method could classify by an accuracy level of 100% at quantization level 8 with pixel distances of 50, 60, and 70. In other words, the SVM method obtained higher accuracy than the KNN method

Cervical cancer is one of the female reproductive health diseases being a significant issue globally because of the large number of new cases and deaths, particularly among women in developing countries. Cervical cancer can be avoided if detected early. The Pap smear screening procedure is used in industrialized nations to detect cervical cancer early. However, limited human resources, a significant time commitment, high prices, and insufficient infrastructure make it less successful in developing countries. With three types of cervical cell images: Normal, Low-grade Squamous Intraepithelial Lesion (LSIL), and High-grade Squamous Intraepithelial Lesion (HSIL), this study offers a classification system for cervical cell images using an image processing technique called Gray Level Co-occurrence Matrix (GLCM) and a Support Vector Machine (SVM) classification method (HSIL). With HSIL class as positive data and LSIL and Normal as negative data, the classification system used three SVM models: Cubic, Quadratic, and Fine Gaussian. SVM classification accuracy was 97.5 percent for 3.54s using the GLCM feature extraction approach.

This paper proposes an optimal control strategy for a standalone wind energy system. In Indonesia, the use of renewable energy, especially wind turbines, is increasingly popular because of the potential winds are very prospective. Based on the existing wind potential, generally, the wind turbine used in Indonesia is on a small scale. This small scale wind turbine is best suited to use a permanent magnet synchronous generator (PMSG) type generator. In operation, the wind turbine system implements the off-grid system to the distribution network because it adjusts the highly volatile wind speed. In order to improve the performance of this wind turbine system, then in this research conducted a strategy based on maximum power point tracking (MPPT) control. The MPPT control-based control strategy is applied to converters connected to PMSG generators. This MPPT control strategy is supported by perturbing and observe (PO) algorithm. In this study, a performance test on a 1000 watt wind turbine system operating at wind speeds ranges from 3 to 10 m/s is done. This performance test is simulated in Simulink-Matlab software. The results showed that the use of MPPT control on wind turbine systems capable of increasing the power output significantly.

Recent studies are directed towards achieving the goal of buildings with a net energy of zero. In Iraq, the price of energy units (kwhr) for government buildings is high and not supported. In this paper, a case study is conducted for one of these buildings (the building of the college of electronics engineering located in Mosul city north of Iraq) to reach to net-zero energy building (NZEB). The available renewable energy resources are proposed represented by solar energy. HOMER (Hybrid Optimization of Multiple Energy Resources) software is used to find the size of the grid-connected photovoltaic system to reach NZEB. The energy consumed by the building is recorded hourly for one year and the cost of the photovoltaic system and the inverter are obtained as local prices. With the help of smart meters which is have the ability to measure the bought energy from the grid and sell energy to the grid. The result showed that using 173 panels in the roof of the proposed building with an installed capacity of 71.795kw and an inverter with 75kw lead makes the cost of energy is zero over the project lifetime.

The limitations of fossil fuels make renewable energy power plants increasingly popular. The power plant is usually integrated into an electric power distribution network called a distributed generation (DG). The integration of DG in the distribution network makes the network scheme change. We need to do some re-planning with the presence of DG to improve distribution network performance. This paper discusses applying the genetic algorithm (GA) method for optimization to improve distribution network performance. The presence of DG in the distribution network makes the system more dynamic. The GA method with the ability to avoid local minima is the answer to the existing problems. The system test was carried out on an IEEE 69-bus model distribution network. The results showed that the GA method was able to produce distribution network optimization with a significant reduction in power losses while at the same time increasing the quality of the bus voltage.

In this paper, it is developed an analysis of the usage of meta-heuristic models inspired by nature to obtain the Optimal Reactive Power Dispatch of a 30 bus system. Firstly, the objective function is presented, considering its power flow parameters. Afterward, the main algorithms were introduced the Grey Wolf Optimization (GWO), its improved version (I-GWO), and the Whale Optimization Algorithm (WOA). Then, they were implemented on the IEEE 30 bus system to calculate the objective function with 13 decision variables and 7 restrictions. Finally, the main results were presented, where it is highlighted an improvement in the Objective Function of 0.45%, in comparison with the traditional Optimal Power Flow implemented in Matpower.

This research integrates the dynamic line rating system into an optimal power flow problem. The application is to calculate the line rating using the heat balance equation based on its temperature rating. As a result, line rating that is usually static can be dynamic that makes the value higher so the system would be better. But because of this, the problem of optimal power flow becomes more difficult because of the addition of temperature variable control and heat balance equation constraint that are quite complex. The modelling also requires considering transmission losses into the Direct Current Optimal Power Flow (DCOPF) problem. This study uses a quadratically constrained quadratic program method in a direct current optimal power flow problem considering transmission losses with the integration of dynamic line rating based on the heat balance equation. Testing method are performed on IEEE 9 bus and 30 bus system. The results show that the method used can accommodate the heat balance equation and show that the integration of dynamic line rating makes the system more economical.

One of the 5G key features in improving energy efficiency is suitable for cognitive radio technology. Cognitive radio has the ability in sensing the operation situation and providing a certain response regarding the sensing result. The existence of multipath due to many obstacle during the propagation of electromagnetic waves can cause depolarization of the antenna. In this paper, polarization reconfigurable antenna was proposed to support cognitive radio technology in 3.5 GHz 5G system. The antenna consists of a square patch with two slots crossing in the diagonal direction and four switch components in the slot for reconfigure linear polarization, right-hand circular polarization (RHCP), and left-hand circular polarization (LHCP) manually. The results show that antenna has good performance in the perspective of matching impedance and axial ratio. It is shown in simulation that the 𝑺𝟏𝟏 less than -10 dB for RHCP and LHCP with axial ratio 1.762 dB and 1.811 dB respectively. Whereas for linear polarization, the 𝑺𝟏𝟏 of 3.5 GHz more than -10 dB and has optimum value at 3.1 GHz with an axial ratio of 40 dB.

Through the wall radar (TWR) is a technology to detect the presence of an object behind a wall. Recently, the need to detect multiple objects simultaneously has driven an improvement for this technology. A problem is coming when the distance between two or more objects is too close. They may interfere with each other and decrease its detection accuracy. To improve the detection accuracy, Multiple-Input Multiple-Output (MIMO) method with the finite-difference time-domain (FDTD) modeling for 2-dimensional TWR is presented. A pair of simulation schemes, Single Input Single Output (SISO) and MIMO, has been done to examine this method's performance by changing the object's gap. It compares the B-Scan result of three schemes. This result shows that MIMO can increase the detection resolution and to separate two objects which are very close to each other despite they have different SNR. Furthermore, if both objects' SNR are similar, the MIMO detection ability is going better when compared to SISO.

This paper proposes a study on UTHM's shuttle bus services for the UTHM students to get to their desired destination in time. Hence, students refuse to ride the shuttle bus as sometimes the shuttle bus is late and not up to the schedule. Although the university has handed out the shuttle bus schedule through the UTHM KATSANA website, it is still inconsistent due to the traffic congestion. The proposed system required a student to request the services to the selected destination; thus, this application estimates the bus arrival time by locating the bus's live location, route, and bus number. Based on the estimated information, students can plan their ride. Additionally, this application allows the driver to minimise the path length taken to each pit stop by interacting with the students. It also enables them to react and decide before the bus's arrival or wait at the bus stop. Google's distance matrix; Application Programming Interface (API), helps coordinate the chosen bus location to the selected bus stop. There are a total of three development phases in this project. The first phase is database acquisition or collection of existing data related to this project. The second phase is the system's design which involves the Firebase Real-time Database, Google API and Android Studio application. Next, the third phase is the system's testing phase. As a result, the proposed system is simulated and tested in real applications. Finally, the driver's and student's information is recorded in Firebase Real-time Database.

Getting connected with acceptable performance is always a user expectation. Moreover, it is critical when network users are in a mobile position. Receiving an adequate level of signals ensures the users get a good signal to noise ratio (SNR) which then promotes good data connections. Furthermore, measuring the signal from a user perception is always important for optimizing the network coverage. This work proposes a scanning process in a range of frequency to asses the coverage. By the use of software-defined radio (SDR), the research has performed measurement in LTE frequency band 3 in a defined track in the City of Yogyakarta. With a low-cost device, some valuable information such as the actual bandwidth used and the benchmark between operators can be obtained. For instance, the measurement acquired a result that an operator leads covering the track by around 5dB.

The process of feeding laying hens in large quantities and on time manually will experience difficulties. The manual feeding method makes feeding less effective and efficient. A real-time clock (RTC)-based automated feeder for laying hens has been designed in this research. The purpose of designing a laying hens feeder is to assist breeders in automatic feeding management. This tool works by automatically feeding laying hens, which will be active at predetermined time intervals with Arduino Uno as the main system controller. Meanwhile, the RTC provides the time reading by Arduino Uno displayed on the LCD. Then, the Stepper motor functions as a feeding collection actuator and the Servo motor as a feeding valve actuator. This study revealed that the error value of one-week feeding in three feeding slots for three laying hens was 1.66% and in two feeding slots for two laying hens was 0.86%.

The pandemic requires everyone to do regular health checks. Accordingly, it raises concern for anyone because of the fear of being contracted to the covid-19 virus. Every medical service first checks blood pressure because it measures how strong the heart is to pump blood throughout the body. The measurement aims to determine blood pressure levels to prevent various diseases that may affect the patient's condition. Tensimeters, in general, cannot be read by blind patients, and officers in charge simply report their examination results. Therefore, the experiment intended o design a voice-based blood pressure measurement tool for visually impaired patients. Thereupon, the study used the Linear Sequential Model (LSM) method, and the results met expectations and in accordance with the calibration results like a sphygmomanometer in general. Consequently, a digital sphygmomanometer was designed with a sound output based on the ATmega328 microcontroller and can be used by hospitals or self-measurement at home. This tool has an almost similar working principle as any ordinary digital sphygmomanometer using the MPX5700GP pressure sensor, which can measure pressure up to 700psi. The results are displayed on the LCD and voiced through the DFPlayer module containing sound recordings saved on a micro SD.

This paper proposed a modified 555 timer IC with only two comparators by eliminating the discharge transistor and SR flip-flop. A periodic square waveform can be generated by utilizing the interconnection of two comparators to provide 'Set' and 'Reset' conditions to each other. The simulation and experimental results confirm that the clock frequency generated by the modified 555 timer circuit is significantly higher and duty cycle adjustment can be made more precise and flexible than the commercially available 555 timer circuit. A frequency in the band of 1MHz to 89MHz can be generated by replacing the resistors, capacitor, and comparator specification in all possible configurations. As a result, a lower cost and higher efficiency on-chip reference clock generator can be realized to fulfill versatile future application needs.

This paper proposes an adaptive voltage DC power supply design to overcome the voltage sag problem that occurs when starting an induction motor. When an induction motor is started, the inrush current drawn by the motor causes a decrease in the voltage across the grid-connected to this motor. This voltage sag can be compensated by injecting voltage into the grid. Because the voltage drop fluctuates, the injection voltage also needs to be adjusted in magnitude. Instead of using duty cycle control to obtain the appropriate voltage magnitude required, this paper proposes to use a variable resistor (VR) combination to control the output voltage of a boost converter. In this preliminary study, the voltage drop is divided into four regions, with different compensation voltage magnitudes for each region. The magnitude of the injected voltage during the voltage sag will be adjusted according to the magnitude of the voltage drop that occurs. The results of this study indicate that the voltage magnitude required to be injected into the grid can be met appropriately. The selection of the VR resistance value according to the magnitude of the voltage drop provides the appropriate injection voltage value as well. Further research is needed to increase the voltage area to be injected by the voltage from this power supply so that the voltage difference between the grid and the power supply can be reduced and produce a smaller voltage ripple.

Low current ripple in the boost converter is crucial in renewable energy system design due to its potential impacts on the operation of other devices. One device that can be directly affected is the MPPT controller. To address this issue, a modified boost converter has been introduced to yield a lower current ripple compared to that of the standard boost converter. This paper presents a study conducted on the modified boost converter when connected to a PV array and a switch controlled by MPPT utilizing Perturb & Observe (P&O) and Particle Swarm Optimization (PSO) algorithms. The simulation results show that the PSO algorithm is more reliable than the P&O algorithm under various irradiation conditions. PSO can minimize steady-state oscillation and the capability to track maximum power than P&O.

Many researchers have developed disaster robots. This paper aims to create disaster robot assistant for helping humans in handling victims of natural disasters. This preliminary research is devoted to the interface and instrumentation of the Brushless Direct Current (BLDC) motor. The focus of this paper is on programming techniques for BLDC motors based on Arduino and Digital to Analog Converter (DAC) which are open loop. The robot used to monitor areas that are declared dangerous and can cause natural disasters. In this preliminary research, an assist robot was created that runs on roads that are not intended to help humans carry several environmental monitoring devices on volcanoes. The robot carries an Unmanned Aircraft Vehicle (UAV) equipped with several sensors for monitoring environmental conditions. System testing is to provide setpoint data and take measurements of the response in the form of output voltage from the DAC and motor rotation speed (RPM). The measurement results for motor speed (RPM) is 11.915 and for DAC (volt) is 0.0169, where the data is a direct comparison of measurements to measurements from settings using Arduino

This research is about developing noise cancellation system to tackle the problem of hearing loss in industry. Hearing aid may help one to hear sounds, but it will not filter or eliminate background sound. Hence, it is hard to hear clearly. Other than that, for a worker who is working in a high noise environment, repeated exposure to loud noise can lead to severe hearing degradation or permanent hearing loss. There are four proposed filters used for the preprocessing method, and also four proposed adaptive algorithms in the Assistive Listening System (ALS) and as significant finding high Signal-to-Noise Ratio (SNR) and low noise level. Preprocessing method is crucial to minimize the background noises and to eliminate the risk of corrupt hearing. After the findings, the best filter used for preprocessing is Butterworth Lowpass Filter; and the best adaptive algorithm is Smart Noise Canceller. With these procedures, the ALS is able to reduce the noise.

Anatomy learning in medical education has been considered a fundamental foundation. The use of cadavers in anatomy studies has been considered the gold standard. However, due to several factors, the use of cadavers in modern medical education was considered irrelevant. As a result, anatomy learning began to involve the use of technology. One of them is using a virtual reality system. Anatomy learning using the virtual reality system relies on the visual field in presenting the information. Many of the bones make up the structure of the human skeleton that we use as our dataset, which raises the problem of visual attention in presenting content in virtual reality systems. For that, we propose a selective visual attention method (a feature that can isolate objects) to improve the user experience. In this study, we develop an experimental design to investigate the effect of our proposed methods on user experience. The developed design offers credible investigative results. To ensure the accuracy of data collection and the accuracy of data analysis from the developed experimental design, we use methods that suit the research needs. The results of this study are in the form of a planned experimental design that can increase readers' confidence in the results of investigations related to user experience.

Early prevention of heart disease can be done by monitoring the condition of heart's health using an electrocardiograph (ECG). In order to make it easier for medical personnel to monitor heart signals, a media for storing results is needed when an examination is carried out. Storage media will be even more effective if it can be accessed remotely. The purpose of this study was to design a modified ECG to check the condition of the heart using a data logger system. The device is designed in simple and applicable which smartphone acts as the display in real-time and can store the readings. This problem is limited by the research device using one lead to monitor heart signals. This study uses AD8232 as a heart signal receiver from the body and Wemos D1 Mini as a link to WiFi to display data on a smartphone using a created application. The measurement and testing methods carried out are by comparing the heart rate (BPM) value with the Phantom ECG input read on a smartphone, comparing the readings of the human heart signal on smartphone and ECG, and testing the human BPM value on smartphone. The overall results of reading the BPM value in the design obtained the error value under 5%. Sending data from Wemos D1 Mini as a media hotspot programmed using Arduino IDE software works well. The results of the heart signal data that are read on the smartphone can be stored so that users can send or print them.

The Covid-19 that hit the world had an impact on the economy, especially in the trade sector, one of which was experienced by Small and Medium Enterprises (SMEs). Hanura Takeaway (Haway) is an SME engaged in the delivery of goods and food. To facilitate transactions for goods and food delivery services, it is necessary to develop applications that simplify the transaction process. In developing web services, it is necessary to exchange data that is accessed via standard internet protocols. Therefore, we need a web service in developing this application. Implementing a RESTful API web service will certainly facilitate the development of software applications outside the system or with different programming languages or platforms. This research will develop web service architecture using RESTful API in Takeaway application. To optimize the URI, several parameters are used, including filtering, sorting, selection and pagination. The Takeaway application consists of a website as a backend and an Android-based as a frontend. From the test results based on the function method using the Postman application, it shows that the REST API Sever built on the server has been running well. In testing the response time using the Apache JMeter application, the application shows a good response time. Meanwhile, the comparison of responses and requests to SOAP and REST architectures shows that REST takes faster time.

Stroke is one of the leading causes of death and disability in the world. Therefore, it is necessary to diagnose stroke at an early stage and provide an accurate prognostic assessment. This study attempts to classify the severity of stroke based on EEG signals by using statistical parameters of time domain features. The results of this study are expected to diagnose the severity of stroke from the parameters used in the time domain and make decisions about the next treatment steps. In this study, the EEG data was obtained from measurement to stroke patients in public hospital in the city of Kediri. From the EEG, 3 statistical features such as Mean Absolute Value (MAV), Standard Deviation (STD) and Variance were calculated. Stroke severity classes were defined as severe, moderate, and mild. The analyzed EEG frequency sub-bands were Alpha Low (8-9 Hz), Alpha High (9-13 Hz), Beta Low (13-17), and Beta High (17-30 Hz). The label for stroke severity classification as a ground truth uses the NIHSS scale which is assessed by doctors based on visual observations. The results showed that stroke severity classification can be identified by using statistical feature such as MAV, STD and Variance, with EEG sub-band frequency are Alpha Low and Alpha High for grasp motion, Beta Low and Beta High for Elbow motion, and Alpha High and Beta High for shoulder motion. This result showed the potential of using this information as a basic for determining the patient-specific rehabilitation program in the future.

Identification analysis of the malaria parasite cell infection, there is a possibility of human error factor done by paramedics because of the number of samples that must be analyzed. This case is because the human eye tends to be tired while working continuously, which can lead to misclassification and treatment that is not right. Therefore, it takes a computer-based system that facilitates medical expert or laboratory technician in identifying two types of parasite cells namely Plasmodium skizon and Plasmodium gametocytes to reduce instances of human error. This research will be conducted on computer-based identification by processing the image type of plasmodium malariae consists of two types, namely Plasmodium skizon and Plasmodium gametocytes levels using convolutional neural network with VGG-16 pre-trained model using 13 layers and 2 dense layers. This study applied 5-fold cross validation for datasets and the datasets are tested using 4 level epoch nodes. The results showed the success of the classification results which have highest training accuracy 90% as well as the results of the highest testing accuracy 100%. It showed the classification using CNN VGG-16 pre-trained model successfully classified the malaria type images.

The development of Village-Owned Enterprises requires careful planning and preparation from all stakeholders. This study seeks to develop Village-Owned Enterprises based on the information system (IS) development to obtain accounting business process standards agreed by stakeholders. The results can be used as the basis for developing an accounting information system (AIS). The research method used the ADM cycle from TOGAF combined with the accounting cycle. The activity began with the analysis of the preparation stage carried out by Village-Owned Enterprises to reach the maturity level in managing finances. Business process standards approved by Village-Owned Enterprises stakeholders become the basis for developing an accounting business process model oriented towards future business goals as a manifestation of lower-level architecture principles.

Innovation Monitoring chamber temperature, skin temperature, humidity and noise levels in the incubator using the android application is an incubator monitoring tool that can be carried out remotely by utilizing an internet connection to make it easier for medical personnel to monitor continuously and can prevent incubator parameter values from being outside tolerance value. This study aims to monitor chamber temperature, skin temperature, humidity, and noise from the incubator to prevent temperatures that are too hot, humidity that is too high or too low, and high noise levels that can damage baby's hearing by utilizing the internet system (IoT), so that with this tool medical personnel can monitor parameters on android smartphones. This study uses the ESP32 as the mainboard, the Ds18b20 sensor as the skin sensor, the SHT11 sensor as the chamber temperature and humidity sensor, and the analog sound level meter as the noise sensor of the device. From the test results, the largest error at chamber temperature is 0.97% at 36 ̊C setting. In the noise test results, the largest error is 0.73% at setting 55 dB. In testing the skin temperature on the incubator, a value that is close to linear is obtained where when the chamber temperature setting on the tool is increased, the value of the skin temperature on the incubator and tool will also increase. On the results of the humidity test at the setting of 32-37 ̊C, the humidity value of the instrument and calibrator is 68-70%. In the data transmission test, the device can be monitored remotely on the android application as long as the device and smartphone are connected to the internet.

As a private university, Abdurrab University has issued a policy to inspect the performance of the lecturers periodically. To contribute to the implementation of lecturer performance, a computer-based monitoring system is required to meet the policy requirement and should be accessed from both lecturers or the management anywhere and anytime. The computer-based monitoring system has been completely built and currently entering the beta testing phase, where the users are actively involved to provide an assessment based on four exogenous variables and one endogenous variable. By using structural equation modeling (SEM), the results of beta testing have shown that among the fit index, six of them at good fit results, and two others are in the marginal category. This outcome indicates that the system built is feasible to use and acceptable to users.

Orchids are plants that are hard to live in and require extraordinary dealing with for the development cycle. On account of this issue, numerous wild orchids in the timberlands of Indonesia are compromised with annihilation. Orchids are hard to fill in sweltering and dry environments on the grounds that most orchids utilize their underlying foundations to discover food. The solution to this problem is a Smart greenhouse. It is a mini-greenhouse system that can maintain the temperature and humidity of the orchid following its natural ecosystem to make it easier for orchid cultivators to care for the plant. The products work automatically so users do not need to operate or set them manually. It works automatically by adjusting the temperature and humidity according to their natural ecosystem. Users only need to insert orchid seeds or mature orchids into the incubator. Not only that, but users can also monitor the conditions in the orchid growing media through a website-based monitoring system. The results of the average error value of the defuzzification process and the length of time the actuator response turns on based on the defuzzification process. The average error value of the temperature parameter is 1.60%, the average error value of the humidity parameter is 1.22% and the average error value of the defuzzification parameter is 0%. Meanwhile, the average error value of the actuator response time based on the defuzzification process is 0.029%

Nowadays, almost half of the world's population is exposed to air pollution. Sources of pollutants come from energy production, households, industry, transportation, and so on. One of the targets of the sustainable development goals agenda is to pay special attention to air quality in order to obtain the comfort of residents in their daily activities. To determine the level of air pollution in a certain place, an instrumentation system is needed that can measure the levels of pollutants in that place. This paper describes the development of a low-cost air quality measurement system using the MQ7 and MQ135 gas sensors. The detected gases consist of carbon monoxide (CO), carbon dioxide (CO2), and ammonia (NH3). The information from the sensor is then processed by the NodeMCU ESP8266 module which can then be accessed via a certain internet-of-things platform, namely ThingSpeak, and a cloud-based messaging service, namely Telegram. Experimental results show that the concentration of pollutants in a certain place can be known at any time from the ThingSpeak dashboard. In addition, an alert system is created to notify the user about the number of pollutant concentrations and the effects of the presence of these pollutant gases.