cover
Contact Name
Anik Vega Vitianingsih
Contact Email
vega@unitomo.ac.id
Phone
+6281332765765
Journal Mail Official
ijair@unitomo.ac.id
Editorial Address
Jl. Semolowaru no 84, Surabaya, 60118
Location
Kota surabaya,
Jawa timur
INDONESIA
International Journal of Artificial Intelligence and Robotics (IJAIR)
ISSN : -     EISSN : 26866269     DOI : 10.25139
International Journal of Artificial Intelligence & Robotics (IJAIR) is One of the journals published by Informatics Department, Universitas Dr Soetomo, was established in November 2019. IJAIR a double-blind peer-reviewed journal, the aim of this journal is to publish high-quality articles dedicated to the field of information and communication technology, Published 2 times a year in November and May. Focus and Scope: Machine Learning & Soft Computing, Data Mining & Big Data, Computer Vision & Pattern Recognition dan Robotics.
Articles 75 Documents
Automation System for the Disposal of Feces and Urine in Rabbit Cages Using Arduino Yudi Kristyawan; Atinus Yikwa
International Journal of Artificial Intelligence & Robotics (IJAIR) Vol. 3 No. 2 (2021): November 2021
Publisher : Informatics Department-Universitas Dr. Soetomo

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (383.054 KB) | DOI: 10.25139/ijair.v3i2.3347

Abstract

Farming rabbits in large numbers produce large amounts of feces and urine. A cage full of feces and urine can cause health problems for rabbits. This research aims to produce an automation system for the disposal of feces and urine in rabbit cages using an Arduino board and implemented in a prototype form. This system uses electronic devices including load cells, HX711 Module to make it easier to read load cells in measuring weight, real-time clock (RTC) for timing, ultrasonic sensor HC-SR04 to detect the presence of a certain object, dc motor, L298N motor driver module to control a dc motor, an LCD 16x2 module to display the weight and height of feces and urine, a buzzer as a notification of the status of the container if it is full, and an Arduino Uno as a controller of the entire system. The system operates so that the feces excreted by the rabbit fall onto the conveyor belt. At the same time, the urine passes via the conveyor belt and falls into the cross-section before being pumped into the urine collection container. The feces on the conveyor belt will be moved with a dc motor towards the stool container based on a certain time. Each stool and urine container is weighed with a load cell and ultrasonic sensor to detect when the container is full. Then the condition of the load cell and the ultrasonic sensor is displayed on an LCD 16x2. When one or both containers are full, a buzzer will sound as a notification. The method used in this research is an experimental method by manipulating or controlling natural situations into artificial conditions. The artificial condition is the provision of deliberate control over the object of study. The test results show that this system can remove waste based on the time using a belt conveyor and monitoring the weight and height of the dirt. If the dirt has met the specified limit, the system can activate an alarm as a notification.
Recognition of Korean Alphabet (Hangul) Handwriting into Latin Characters Using Backpropagation Method Anang Aris Widodo; Muchammad Yuska Izza Mahendra; Mohammad Zoqi Sarwani
International Journal of Artificial Intelligence & Robotics (IJAIR) Vol. 3 No. 2 (2021): November 2021
Publisher : Informatics Department-Universitas Dr. Soetomo

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (531.095 KB) | DOI: 10.25139/ijair.v3i2.4210

Abstract

The popularity of Korean culture today attracts many people to learn everything about Korea, especially in learning the Korean language. To learn Korean, you must first know Korean letters (Hangul), which are non-Latin characters. Therefore, a digital approach is needed to recognize handwritten Korean (Hangul) words easily. Handwritten character recognition has a vital role in pattern recognition and image processing for handwritten Character Recognition (HCR). The backpropagation method trains the network to balance the network's ability to recognize the patterns used during training and the network's ability to respond correctly to input patterns that are similar but not the same as the patterns used during training. This principle is used for character recognition of Korean characters (Hangul), a sub-topic in fairly complex pattern recognition. The results of the calculation of the backpropagation artificial neural network with MATLAB in this study have succeeded in identifying 576 image training data and 384 Korean letter testing data (Hangul) quite well and obtaining a percentage result of 80.83% with an accuracy rate of all data testing carried out on letters. Korean (Hangul).
Optimization of Breadth-First Search Algorithm for Path Solutions in Mazyin Games Bonifacius Vicky Indriyono; Widyatmoko
International Journal of Artificial Intelligence & Robotics (IJAIR) Vol. 3 No. 2 (2021): November 2021
Publisher : Informatics Department-Universitas Dr. Soetomo

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (381.647 KB) | DOI: 10.25139/ijair.v3i2.4256

Abstract

A game containing elements of artificial intelligence, of course, requires an algorithm in its application. One example of a game that includes elements of artificial intelligence is the Labyrinth game. Maze is a simple educational game. This game is known as finding a way out of the maze to arrive at a predetermined goal. The labyrinth encounters numerous obstacles along the way, such as dead ends and parapets, to reach the target location. In this game, players are required to think logically about how to find the right maze path. The obstacle faced in this game is that sometimes players have difficulty finding a way out, especially if the game level has reached a high level in the process of finding a way out. To solve this problem, a graph tracing technique is needed. The Breadth-First Search (BFS) strategy can be used in conjunction with various graph search algorithms. An example of a broad search method is the Breadth-First Search Algorithm, which works by visiting nodes at level n first before moving on to nodes at level n+1. The advantage of the Breadth-First Search algorithm is that it can find a solution as the shortest path and find the minimum solution if there is more than one solution. This study will discuss how to find a path for the Labyrinth using the BFS algorithm. The result of applying this BFS algorithm is the shortest route solution raised so that the Labyrinth can arrive at the destination point through the route provided.
Comparison of Memetic Algorithm and Genetic Algorithm on Nurse Picket Scheduling at Public Health Center Nico Nico; Novrido Charibaldi; Yuli Fauziah
International Journal of Artificial Intelligence & Robotics (IJAIR) Vol. 4 No. 1 (2022): May 2022
Publisher : Informatics Department-Universitas Dr. Soetomo

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (652.721 KB) | DOI: 10.25139/ijair.v4i1.4323

Abstract

One of the most significant aspects of the working world is the concept of a picket schedule. It is difficult for the scheduler to make an archive since there are frequently many issues with the picket schedule. These issues include schedule clashes, requests for leave, and trading schedules. Evolutionary algorithms have been successful in solving a wide variety of scheduling issues. Evolutionary algorithms are very susceptible to data convergence. But no one has discussed where to start from, where the data converges from making schedules using evolutionary algorithms. The best algorithms among evolutionary algorithms for scheduling are genetic algorithms and memetics algorithms. When it comes to the two algorithms, using genetic algorithms or memetics algorithms may not always offer the optimum outcomes in every situation. Therefore, it is necessary to compare the genetic algorithm and the algorithm's memetic algorithm to determine which one is suitable for the nurse picket schedule. From the results of this study, the memetic algorithm is better than the genetic algorithm in making picket schedules. The memetic algorithm with a population of 10000 and a generation of 5000 does not produce convergent data. While for the genetic algorithm, when the population is 5000 and the generation is 50, the data convergence starts. For accuracy, the memetic algorithm violates only 24 of the 124 existing constraints (80,645%). The genetic algorithm violates 27 of the 124 constraints (78,225%). The average runtime used to generate optimal data using the memetic algorithm takes 20.935592 seconds. For the genetic algorithm, it takes longer, as much as 53.951508 seconds.
Cargo Simulation Robot Prototype with Bluetooth Based Motor Driver Shield Using Arduino Uno Microcontroller Yudi Irawan Chandra; Irfan; Anargya Satya Rifisyah Putro
International Journal of Artificial Intelligence & Robotics (IJAIR) Vol. 4 No. 1 (2022): May 2022
Publisher : Informatics Department-Universitas Dr. Soetomo

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (648.439 KB) | DOI: 10.25139/ijair.v4i1.4326

Abstract

From the android developer website, it can be seen that the number of android users in the world is increasing. Almost 70% of the world's people use Android as their gadget. Nevertheless, today's average person uses android gadgets only to send messages, social media, and telephone. Furthermore, they do not realize they can increase the ease and sophistication of other things in the world and are very useful for everyday things and certain things in life. In today's modern era, many communication types of equipment have an intelligent system or what is commonly referred to as a smartphone. Modern society uses branded gadgets as a lifestyle. Today, the average person uses android gadgets only to send messages, social media, and telephone. When using Bluetooth, applications are generally used to exchange data. However, now Bluetooth is not only used to communicate with telephones or computers but can also be used to command an electronic device according to the needs of its users. The purpose of this research is to create a robot simulation that can be controlled using Bluetooth to move an item to facilitate human work, relieve heavy tasks that have a high risk, for example carrying goods in the factory and reduce accidents in terms of carrying goods, and able to be controlled things remotely as desired by utilizing Bluetooth media using the Arduino Uno-based L293D Motor Shield Driver.
Design of MPPT PV using Particle Swarm Optimization Algorithm under Partial Shading Condition Efendi S Wirateruna; Annisa Fitri Ayu Millenia
International Journal of Artificial Intelligence & Robotics (IJAIR) Vol. 4 No. 1 (2022): May 2022
Publisher : Informatics Department-Universitas Dr. Soetomo

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (411.745 KB) | DOI: 10.25139/ijair.v4i1.4327

Abstract

Fossil energy sources experience a decrease each year when the demand increases significantly. In the case of environmental issues, renewable energy sources (RES) can be energy alternatives. The photovoltaic module is RES with unique characteristics, especially partial shading conditions. This condition leads to the PV characteristic curve experiencing multiple peaks. The paper conducted the simulation of the PV solar panel module using MATLAB Simulink. The Maximum Power Point Tracking (MPPT) PV is also described based on a particle swarm optimization (PSO) algorithm. The proposed algorithm can address multiple peak curve problems due to partial shading conditions. For comparison, the conventional algorithm, perturb & observe, is presented. The PV module is divided into three group cells with irradiance differences for each group to illustrate the partial shading condition. The result shows that the PSO algorithm guarantees optimal and fast response for the operating PowerPoint. It needs about 0.04 seconds to maintain at the optimal power point, 129 Watt, compared with the perturb and observe algorithm performance that only kept at the lower operating power point, 67 Watt at 0.06 second. Thus, the PSO algorithm can tackle the partial shading condition with a fast response to maintain the maximum PowerPoint. Therefore, the PSO algorithm is the proper solution for tracking the optimum operating power point under partial shading conditions.
Optimization of the Number of Clusters of the K-Means Method in Grouping Egg Production Data in Indonesia Solikhun Solikhun; Verdi Yasin; Donni Nasution
International Journal of Artificial Intelligence & Robotics (IJAIR) Vol. 4 No. 1 (2022): May 2022
Publisher : Informatics Department-Universitas Dr. Soetomo

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (355.569 KB) | DOI: 10.25139/ijair.v4i1.4328

Abstract

The need for eggs that continues to increase will not increase with large egg production so that there is a shortage of egg supplies which results in high egg prices. It is necessary to group egg production in Indonesia to find out which areas fall into the high cluster and which areas fall into the low cluster. This study aims to classify the egg production of laying hens in Indonesia. The method used is the K-Means Clustering method which is a popular clustering method. To find out how optimal the number of clusters in the K-Means method is for grouping egg production in Indonesia, the researcher evaluates the DBI value of each number of existing clusters. In this study, 8 clusters were used, namely 2 clusters, 3 clusters, 4 clusters, 5 clusters, 6 clusters, 7 clusters, 8 clusters, and 9 clusters. The results of measuring the DBI value are the number of clusters 2 = 0.215, the number of clusters 3 = 0.149, the number of clusters 4 = 0.146, the number of clusters 5 = 0.157, the number of clusters 6 = 0.180, the number of clusters 7 = 0.205, the number of clusters 8 = 0.192 and the number of clusters 9 = 0.154. This study shows that the best number of clusters is the number of clusters 4 with the smallest DBI value of 0.146.
Optimization of Foodstuffs for Patients with Hypertension Using the Improved Particle Swarm Optimization Method Novrido Charibaldi; Nur Heri Cahyana; Muhamad Setiawan Wicaksono
International Journal of Artificial Intelligence & Robotics (IJAIR) Vol. 4 No. 1 (2022): May 2022
Publisher : Informatics Department-Universitas Dr. Soetomo

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (365.257 KB) | DOI: 10.25139/ijair.v4i1.4335

Abstract

Hypertension is when a person's blood pressure exceeds the reasonable limits determined by experts. A person who suffers from high blood pressure or hypertension risks developing non-communicable diseases that can endanger the sufferer's life, such as stroke and heart attack. One of the causes that can increase and worsen hypertension is an unhealthy lifestyle. Due to a lack of knowledge in regulating food composition, it is difficult for ordinary people to vary the composition of food in the next few days, which is usually done by simply avoiding foods ordered by doctors or experts. The Improved Particle Swarm Optimization (IPSO) method was chosen because it can be used to solve the problem of optimizing optimal food composition. In addition, the IPSO method can also remember the worst position ever visited so that particles can pass through a bad position and always try a better position. Based on the research conducted, the IPSO method succeeded in producing recommendations for the composition of foods consumed by people with hypertension consisting of 3 portions, namely breakfast, lunch, and dinner. Breakfast and lunch contain staple foods, plant sources, animal sources, vegetables, fruits, or complementary foods. At the same time, dinner contains only staple foods, animal sources, plant sources, and vegetables. This research found that the iteration that can produce optimal results is 400 iterations and the most optimal particles are 10 particles. This happens because the price of food ingredients is included in the calculation.
Improvement Of Query Speaking on The Indonesian to Madura Dictionary Using Levenshtein Distance Method M. Yahya Ubaidillah; Muchamad Kurniawan; Septiyawan Rosetya Wardhana
International Journal of Artificial Intelligence & Robotics (IJAIR) Vol. 3 No. 2 (2021): November 2021
Publisher : Informatics Department-Universitas Dr. Soetomo

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (354.531 KB) | DOI: 10.25139/ijair.v3i2.4394

Abstract

Men are distinguished from other living beings by their use of language, which becomes one of their most distinctive and humanistic qualities. Many different languages are spoken worldwide, including Indonesian, which has approximately 742 different dialects. Due to the unique language of Madura, which is located on a large island with numerous beach tourism destinations, tourists will have difficulty navigating the island. People outside Madura Island who come to visit or vacation will find it difficult to communicate with the locals during their stay or holiday. An Indonesian to Madurese translation dictionary is therefore required in this case. The Levenshtein Distance method was employed in this investigation. The algorithm in the dictionary is used to process the search for the closest distance (dif) between the words being inputted and the words that are already in the database. To provide a prototype for the use of dictionaries. Indonesian and Madurese data sets were used in the investigation by the researcher. According to the simulation results acquired after multiple trials, the error accuracy was 90 % for the first letter input, 84 % for the middle letter input, and 84 % for the last letter input for the first letter. As a result, according to the study's findings, the accuracy of this dictionary increased by 86 %. The first letter received 90 % of the votes, the middle letter received 84 %, and the last letter received 84 %. As a result, according to the study's findings, the accuracy of this dictionary increased by 86 %. The first letter received 90 % of the votes, the middle letter received 84 %, and the last letter received 84 %. As a result, according to the study's findings, the accuracy of this dictionary increased by 86 %.
Convolutional Neural Network Method for Classification of Syllables in Javanese Script Yuli Fauziah; Kevin Aprilianta; Heru Cahya Rustamaji
International Journal of Artificial Intelligence & Robotics (IJAIR) Vol. 3 No. 2 (2021): November 2021
Publisher : Informatics Department-Universitas Dr. Soetomo

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (543.978 KB) | DOI: 10.25139/ijair.v3i2.4395

Abstract

Javanese script is one of the languages which are a typical Javanese culture. Javanese script is seen in its use in writing the name of a particular agency or location that has historical and tourism value. The use of Javanese script in public places makes the existence of this script seen by many people, not only by the Javanese people. Some of them have difficulty recognizing the Javanese characters they encounter. One method of pattern recognition and image processing is Convolutional Neural Network (CNN). CNN is a method that uses convolution operations in performing feature extraction on images as a basis for classification. The process consists of initial data processing, classification, and syllable formation. The classification consists of 48 classes covering Javanese script types, namely basic letters (Carakan) and voice-modifying scripts (Sandhangan). It is tested with multi-class confusion matrix scenarios to determine the accuracy, precision, and recall of the built CNN model. The CNN architecture consists of three convolution layers with max-pooling operations. The training configuration includes a learning rate of 0.0001, and the number of filters for each convolution layer is 32, 64, and 128 filters. The dropout value used is 0.5, and the number of neurons in the fully-connected layer is 1,024 neurons. The average performance value of accuracy reached 87.65%, the average precision value was 88.01%, and the average recall value was 87.70%.