Garuda - Garba Rujukan Digital

KLASIFIKASI DAN DETEKSI KERETAKAN PADA TROTOAR MENGGUNAKAN METODE CONVOLUTIONAL NEURAL NETWORK: Classification and Detection of Cracks on Sidewalks Using the Convolutional Neural Network Method ari Wibowo; E Setiyadi
JURNAL TEKNIK SIPIL CENDEKIA (JTSC) Vol 4 No 1 (2023): February
Publisher : Departement of Civil Engineering, Universitas Winaya Mukti

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51988/jtsc.v4i1.116

Trotoar adalah bagian dari jalan raya yang khusus disediakan untuk pejalan kaki dimana trotora pada umumnya terletak di daerah manfaat jalan untuk memudahkan ketika berjalan kaki. Hal ini agar pejalan kaki tidak bercampur dengan kendaraan yang tentunya dapat memperlambat arus lalu lintas dan dapat membahayakan pejalan kaki itu sendiri. Namun pada kenyataannya permukaan trotoar memiliki kondisi yang beragam. Oleh karena itu perbaikan trotoar merupakan solusi tepat agar kerusakan trotoar tidak semakin memburuk dan tidak mengganggu para pengguna trotoar. Langkah pertama dalam permukaan trotoar adalah mendeteksi kerusakan yang ada di permukaan. Salah satu metode yang dapat dipakai dalam mendeteksi kerusakan pada trotoar adalah menggunakan teknologi terkini salah satunya adalah memanfaatkan deep learningdengan metode CNN. Tujuan penelitian ini adalah menyusun algoritma yang secara khusus digunakan untuk membedakan trotoar yang retak dan tidak retak. Adapun dataset latih yang digunakan sebanyak 3200 citra gambar dan 800 citra untuk data uji. Dimana gambar gambar ini kami ambil dari katalog kaggle. Dari penelitian yang kami lakukan hasil pengujian menunjukkan bahwa model berhasil membedakan permukaan trotoar yang retak maupun yang tidak retak dengan akurasi yang cukup tinggi, dimana nilai akurasi rata-rata yang dihasilkan di atas 96% dan nilai loss yang mendekati 1,5%.

Implementasi Algoritma Deep Learning You Only Look Once (YOLOv5) Untuk Deteksi Buah Segar Dan Busuk Lusiana Lusiana; Ari Wibowo; Tika Kartika Dewi
Paspalum: Jurnal Ilmiah Pertanian Vol 11, No 1 (2023)
Publisher : Lembaga Penelitian dan Pengabdian Universitas Winaya Mukti

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35138/paspalum.v11i1.489

Fruit is one of the nutritional needs for the body that must be met. But with a note, these nutrients will be obtained from fruit that is still fresh. The definition of fresh fruit itself is fruit that can be consumed directly and does not require any further processing. There are many ways to select and differentiate between fresh fruit and bad fruit and in general direct observations can be made. But over time, there are several other ways to observe fruit freshness using existing technology. Where one of them is by optimizing deep learning and machine learning. This detection and classification system was created using a deep learning method using the YOLOv5 algorithm which can detect in real-time the types of apples, bananas and oranges. We use image datasets for each of these fruits for fresh fruit and rotten fruit, a total of 1200 images for train data and 330 images for validation data and 6 images for test data. Based on the tests that have been carried out with training data, along with validation data, and test data using the YOLOv5 algorithm, it can be concluded that this detection method can recognize objects consistently with a high degree of accuracy. This can be proven at the level of accuracy which reaches an accuracy rate of 90%.

DETEKSI KERETAKAN JALAN ASPAL MENGGUNAKAN METODE CONVOLUTIONAL NEURAL NETWORK ari Wibowo; yusuf yulianto
JURNAL TEKNIK SIPIL CENDEKIA (JTSC) Vol 4 No 2 (2023): July
Publisher : Departement of Civil Engineering, Universitas Winaya Mukti

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51988/jtsc.v4i2.132

Road conditions determine the comfort of road users, the comfort of these road users is the responsibility of the Public Works and Spatial Planning Office in each region. Roads are of course an important aspect because roads are the main supporting factor in the social, cultural, environmental fields which are developed in order to achieve an equitable distribution of development between regions and sustainability with regional and economic development approaches. The first step that must be taken by policy makers in seeking comfort for users is to evaluate the quality of roads, including in Indonesia. The evaluation in question includes estimating repairs, required construction, estimating quality. The strategic step in making road quality evaluation steps is to detect road cracks on the surface. One of them is by implementing an intelligent system method in detecting road damage using the Convolutional Neural Network (CNN) algorithm. The input is an image of the road surface in RGB format. The image is obtained from kaggle as many as 2074 images. Based on the results of the tests and evaluations that have been carried out, it can be concluded that the system built has succeeded in producing very good data as evidenced by an accuracy rate of 92.9%.

FLUID DYNAMIC SIMULATION ON THE FLARE OF COMBUSTION OF GAS FROM BIOMASS GASIFICATION dian susanto; Muhtar Kosim; Ari Wibowo
Jurnal Mekanika dan Manufaktur Vol 3 No 1 (2023)
Publisher : Universitas Majalengka

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31949/jmm.v3i1.5682

The use of energy which always comes from fossil fuels will eventually run out, so the development of renewable energy or alternative energy is very important to maintain petroleum reserves and as a substitute for fossil fuels which are the main energy source. One alternative energy is biomass which has not been widely used by the gasification method. The gas produced by the gasification process is utilized by burning it in a flare to get a flame. In this study, the 3D simulation method with Computational Fluid Dynamics (CFD) was used to determine the temperature distribution on the flare walls using CFD simulations and to compare the temperature of the flare walls from the CFD simulation results with the test results. The results of this study, the distribution of combustion occurs in the flare with a temperature of 1106°C in the upper area close to the outlet boundary. The wall temperature comparison shows that the CFD simulation tends to be similar to the test results. This shows that computational fluid dynamic simulations can be used to predict fluid flow rates and combustion reactions.

OPTIMIZATION OF PREDICTION AND PREVENTION OF DEFECTS ON METAL BASED ON AI USING VGG16 ARCHITECTURE muhtar kosim; Ari Wibowo; Novandri Tri Setioputro; Kasda; Dian Susanto
Jurnal Mekanika dan Manufaktur Vol 3 No 1 (2023)
Publisher : Universitas Majalengka

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31949/jmm.v3i1.6542

Manufacturing is one of the most valuable industries in the world, it can be automated without limits but still stuck in traditional manual and slow processes. Industry 4.0 is racing to define a new era in digital manufacturing through the implementation of Machine Learning methods. In this era, Machine learning has been widely applied to various fields and will certainly be very good applied in the manufacturing world. One of them is used to predict and prevent defects in metal. The process of predicting and preventing defects in metal is one of the important efforts in improving and maintaining production quality. Accuracy in predicting and preventing defects in metal can be an innovation and competitiveness in technology, both in production methods, and improving product safety and its users. Human operators and inspectors without digital assistance generally can spend a lot of time researching visual data, especially in high-volume production environments. For this reason, there needs to be research in developing Machine Learning technology in an effort to prevent the occurrence of defects in metal. And one of the development of this technology by using Convolutional Neural Network (CNN) architecture Visual Geometry Group 16 layer (VGG16). As for the metal defect dataset with 10 classes with details for training data as many as 17221, and test dataset as many as 4311, From the use of methods and datasets available, has been done training model used and produce very good accuracy, that is equal to 89% and testing with accuracy equal to 76%. And also done Interpreter process against new input data, to know metal defect type, prediction accuracy and appropriate action to prevent and overcome metal defect type result of Interpreter process application.

Implementasi Algoritma Deep Learning You Only Look Once (YOLOv5) Untuk Deteksi Buah Segar Dan Busuk Lusiana Lusiana; Ari Wibowo; Tika Kartika Dewi
Paspalum: Jurnal Ilmiah Pertanian Vol. 11 No. 1 (2023)
Publisher : Lembaga Penelitian dan Pengabdian Universitas Winaya Mukti

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35138/paspalum.v11i1.489

Fruit is one of the nutritional needs for the body that must be met. But with a note, these nutrients will be obtained from fruit that is still fresh. The definition of fresh fruit itself is fruit that can be consumed directly and does not require any further processing. There are many ways to select and differentiate between fresh fruit and bad fruit and in general direct observationsÂ can be made. But over time, there are several other ways to observe fruit freshness using existing technology. Where one of them is by optimizing deep learning and machine learning. This detection and classification system was created using a deep learning method using the YOLOv5 algorithm which can detect in real-time the types of apples, bananas and oranges. We use image datasets for each of these fruits for fresh fruit and rotten fruit, a total of 1200 images for train data and 330 images for validation data and 6 images for test data. Based on the tests that have been carried out with training data, along with validation data, and test data using the YOLOv5 algorithm, it can be concluded that this detection method can recognize objects consistently with a high degree of accuracy. This can be proven at the level of accuracy which reaches an accuracy rate of 90%.

PENELUSURAN JEJAK BANGUNAN KOLONIAL DI INDONESIA BERBASIS FAÃ‡ADE BANGUNAN MENGGUNAKAN METODE CNN ARSITEKTUR VGG16 Susanto Susanto; Ari Wibowo
Jurnal Arsitektur ARCADE Vol 7 No 3 (2023): Jurnal Arsitektur ARCADE September 2023
Publisher : Prodi Arsitektur UNIVERSITAS KEBANGSAAN

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31848/arcade.v7i3.3179

Abstract: Colonialization in Indonesia, especially that carried out by the Dutch, is one of the important historical chapters in Indonesia because it was able to change the way of thinking of architecture in the Dutch East Indies to be more modern, approaching what happened in Western countries. The influence of modernism in architecture can be seen in the faÃ§ade, shape, strength, and other important parts. Broadly speaking, colonial buildings are buildings that have patterns and characteristics that other buildings do not have. Therefore, the building is capable of being recognized. One way to quickly identify and trace the Dutch colonial legacy is to use the CNN method. Using the CNN method shows that the input data used can be recognized very well. Because the level of accuracy in the training stage is 97% and the accuracy value of each input image detection is more than 90%. One example is the Maybank Surabaya building, which includes a colonial architectural style with an accuracy of 100%, the Zeiss telescope dome is included in a colonial architectural style building with an accuracy rate of 99%, and other examples of buildings.Keyword: Kata Kunci 1, Colonial Architecture 2, Architectural style detection 3, CNNAbstrak: Kolonialisasi di Indonesia terutama yang dilakukan oleh Belanda merupakan salah satu babak sejarah penting di Indonesia karena mampu merubah cara berfikir arsitektur di Hindia Belanda semakin modern mendekati yang terjadi di negara Barat. Pengaruh modernism dalam arsitektur tersebut dapat dilihat pada faÃ§ade, bentuk, kekuatan, dan bagian penting lainnya. Secara garis besar bangunan colonial merupakan bangunan yang memiliki corak dan ciri khas yang tidak dimiliki oleh bangunan lain. Oleh karena itu, bangunan tersebut mampu untuk dikenali. Salah satu cara dalam mengenali maupun menelusuri jejak peninggalan colonial belanda secara cepat adalah dengan menggunakan metode CNN. Dengan menggunakan metode CNN menunjukkan bahwa data inputan yang dipakai mampu dikenali dengan sangat baik. Dikarenakan tingkat akurasi di tahap pelatihan sebesar 97% serta nilai akurasi tiap deteksi gambar inputan yang mencapai 90% lebih. Salah satu contohnya Gedung maybank Surabaya termasuk tipe gaya arsitektur colonial dengan akurasi sebesar 100%, kubah teleskop Zeiss termasuk ke dalam bangunan dengan gaya arsitektur colonial dengan tingkat akurasi sebesar 99% dan contoh bangunan lainnya.Kata Kunci: Kata Kunci 1, Arsitektur kolonial Â 2, Deteksi gaya arsitektur 3, CNN

ANALISIS KEEFEKTIFAN PENCAHAYAAN BUATAN DALAM RUANGAN MENGGUNAKAN MACHINE LEARNING (Studi Kasus Data Sensor Lingkungan Dan Konsumsi Listrik Smart Building) Susanto Susanto; Ari Wibowo
Jurnal Arsitektur ARCADE Vol 8 No 1 (2024): Jurnal Arsitektur ARCADE Maret 2024
Publisher : Prodi Arsitektur UNIVERSITAS KEBANGSAAN

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31848/arcade.v8i1.3828

Abstract: Lighting in an office is an important aspect which can certainly support the performance of its use. Especially in artificial lighting, it is not uncommon if the design system does not adapt to natural daylight, of course it is not effective, and has a negative impact and causes losses periodically. Therefore, it is important to know the effectiveness of using artificial lighting in a room. In this analysis process, one method that can be used is machine learning. By using CU-BEMS based data, which is obtained from environmental sensor data and electricity consumption. This data was taken over a period of 18 months, consisting of several readings from equipment installed in each room. And from this data it is used to select the required quantities so that they can suit your needs. After that, the XGBoost algorithm can be applied and the approach process carried out produces very good accuracy. This can be proven from the results of the resulting evaluation matrix, such as the R-Squared value of 0.99, the RMSE (Root Mean Square Error) is 0.028. Then, in the data testing process, high accuracy was also obtained, for R-Squared of 0.99 and RMSE (Root Mean Square Error) of 0.092. After the training and testing process has obtained very good results, we proceed to the process of analyzing the effectiveness of artificial lighting by entering certain input values, to produce information about the effectiveness of artificial lighting in certain rooms, accompanied by the level of confidence or accuracy of each prediction result.Keyword: Artificial Lighting, Sensor data and electricity consumption, Machine Learning, XGBooAbstrak: Pencahayaan pada kantor merupakan salah satu aspek penting yang tentu dapat menunjang kinerja penggunaannya. Khususnya dalam pencahayaan buatan, tidak jarang jika sistem perancangannya tidak menyesuaikan akan pencahayaan alami siang hari, tentulah tidak efektif, dan berdampak negative dan memberikan kerugian secara berkala. Oleh karena itu, penting untuk mengetahui keefektifan dari penggunaan pencahayaan buatan pada suatu ruangan. Dalam proses analisis ini salah satu metode yang dapat digunakan adalah dengan menggunakan machine learning. Dengan menggunakan data berbasis CU-BEMS, yang diperoleh dari data sensor lingkungan dan konsumsi listrik. Data ini diambil dalam kurun waktu selama 18 bulan, yang terdiri beberapa hasil pembacaan alat yang terpasang di tiap ruangan. Dan dari data tersebut digunakan untuk menyeleksi besaran yang diperlukan sehingga dapat sesuai dengan kebutuhan. Setelah itu, maka penerapan algoritma XGBoost dapat dilakukan dan dari proses pendekatan yang dilakukan, menghasilkan akurasi yang sangat baik. Hal ini dapat dibuktikan dari hasil matrik evaluasi yang dihasilkan seperti nilai R-Squared sebesar 0.99, besar RMSE (Root Mean Square Error) yaitu 0.028. KemudianÂ pada proses pengujian data juga diperoleh akurasi yang tinggi, untuk R-Squared sebesar 0.99 dan RMSE (Root Mean Square Error) 0.092. Setelah proses pelatihan dan pengujian didapatkan hasil yang sangat baik, maka dilanjutkan ke proses analisis keefektifan pencahayaan buatan dengan mengisikan nilai inputan tertentu, untuk menghasilkan informasi mengenai keefektifan dari pencahayaan buatan pada ruangan tertentu dengan disertai tingkat kepercayaan atau keakurasian dari setiap hasil prediksinya.Kata Kunci: Pencahayaan Buatan, Data sensor dan konsumsi listrik, Machine Learning, XGBoost.

Integrasi Gravity Model dan Indeks Kansky dalam Evaluasi Struktur Ruang Wilayah Alfa Yoseph; Ari Wibowo; Susanto Susanto
JAUR (JOURNAL OF ARCHITECTURE AND URBANISM RESEARCH) Vol. 10 No. 1 (2026): Oktober (INPRESS)
Publisher : Universitas Medan Area

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31289/jaur.v10i1.18992

Regional connectivity is a fundamental aspect of spatial structure planning because it influences accessibility, spatial interaction among activity centers, and the efficiency of transportation systems. Subang Regency, one of the rapidly developing regions in West Java, has experienced significant economic growth and urban expansion, resulting in disparities in regional connectivity, increasing traffic volume, and suboptimal road network performance. This study aimed to evaluate the regional spatial structure of Subang Regency by integrating the Gravity Model and the Kansky Connectivity Index as a quantitative approach to identify the level of spatial interaction among activity centers and the quality of road network connectivity. A descriptive quantitative method was employed using secondary data, including population, inter-center distances, road network length, administrative area, and road network topology covering five Local Activity Centers (PKL) and twenty-four Area Service Centers (PPK). The analysis consisted of Gravity Model calculations to measure spatial interaction, road density analysis, and the Kansky Index (β) to assess road network connectivity. The integration of the Gravity Model and the Kansky Index provided a more comprehensive framework for evaluating regional spatial structure and could support infrastructure development priorities and sustainable regional planning in Subang Regency.

Title

Found 9 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 9 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 9 Documents
Search