Murman Dwi Prasetio

Dr. Murman Dwi Prasetio

Telkom University, Bandung, Jawa Barat, Indonesia
orcid

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
15 September 2022 in Buildings
Verified
Journal Article
Open access
PDF file
Poh Kiat NgPoh Kiat Ng
Poh Kiat Ng
Murman Dwi PrasetioMurman Dwi Prasetio
Murman Dwi Prasetio
KLKia Wai Liew
Kia Wai Liew
BLBoon Kian Lim
Boon Kian Lim
AOAyudita Oktafiani
Ayudita Oktafiani
YSYunita Nugrahaini Safrudin
Yunita Nugrahaini Safrudin
The Indonesian roof tile manufacturing industry relies heavily on manual operations, specifically in transportation and inspection processes, which creates multiple issues, such as fatigue, injuries, human error, and reduced productivity. Various industries in the Indonesian industrial landscape have begun embracing a problem-solving approach known as the theory of inventive problem-solving (TRIZ) to mine solutions for industrial issues. Nevertheless, its application in the Indonesian roof tile manufacturing industry remains unaddressed. The study aims to solve manual handling issues in the roof tile manufacturing industry using TRIZ. Three observations were outlined from manual roof tile transportation and inspection, followed by the formulation of engineering contradictions (ECs). The ECs were linked with system parameters, which were used as indicators within the contradiction matrix to extract inventive principles as solution models for conceptual development. The concept included an automated system with a conveyor belt (#15: dynamics) for effective transportation, automated image capture (#28: mechanics substitution) for effective inspection, and a flipping conveyor (#25: self-service) to eliminate manual contact. Although the study addressed several issues stemming from manual operations, mechanical analysis, prototyping, and usability testing still require improvements.

14 December 2020
Verified
Conference Paper
Closed access
Murman Dwi PrasetioMurman Dwi Prasetio
Murman Dwi Prasetio
MRMohammad Husain Rifai
Mohammad Husain Rifai
RXRais Yufli Xavierullah
Rais Yufli Xavierullah
The current industry has been developing rapidly which make the company must have high competitiveness by maintaining the product’s quality and quantity produced by the company. PT. XYZ is one of the companies in the clay tile industry. There are classifications of PT. XYZ products: good tile, white stone tile, and cracked tile in quality control. During its classification, PT XYZ still uses the traditional method of vision. The traditional detection of errors or defects only using human vision can slow down the process and increase the error rate. With the rapid development of automation can overcome this by the discovery of artificial visual detectors that use measurement methods, image preprocessing, and algorithms in detecting these defects. In this study using the Support Vector Machine (SVM) method for classifying defects and Local Binary Pattern (LBP) method for extracting the feature on tiles. Direct image taking in this study use raspberry pi and create the algorithm system using python software. The results of this study concluded that the highest level of accuracy was 87.5% using linear kernels. While the required time for direct classification is 10.63 seconds.

1 November 2017
Verified
Conference Paper
Closed access
Murman Dwi PrasetioMurman Dwi Prasetio
Murman Dwi Prasetio
THTomohiro Hayashida
Tomohiro Hayashida
INIchiro Nishizaki
Ichiro Nishizaki
SSShinya Sekizaki
Shinya Sekizaki
The paradigm theory of neural network (NN) is expressing from biological of human brain function system. In this study, we deal with the deep belief network (DBN) model using the speech as natural language processing for training the system. The characteristic of a deep belief network is similar to a neural network. Whereas the performance of a neural network depends on the structure itself and it is suitable to select the model and size of the network for the data to handle. In the occasion, deep belief network has the advantage in speech recognition it generates the feature learning with a subsequent stage of supervised learning, once the network initialized, the first way using unsupervised and then fine-tuned with the labeled data to train some neuron in initial weight vectors. Deep belief network (DBN) have many nonlinear hidden layers to produce posterior of probabilities that take several frames of coefficient as input. This paper shows the good performance improvement of modularity in DBN as combination technique of computation.

1 September 2021 in International Journal of Industrial Optimization
Verified
Journal Article
Open access
PDF file
Murman Dwi PrasetioMurman Dwi Prasetio
Murman Dwi Prasetio
RXRais Yufli Xavier
Rais Yufli Xavier
HRHaris Rachmat
Haris Rachmat
WWWiyono Wiyono
Wiyono Wiyono
DADenny Sukma Eka Atmaja
Denny Sukma Eka Atmaja
The strength of the company's competitiveness is needed because the current industrial development is very rapid. It is necessary to maintain the quality and quantity of the products produced according to company standards. One of the companies that must maintain the quality and quantity is PT. XYZ is a clay tile company. The classification of products used by this company to maintain good quality is three classes: good tile, white stone tile, and cracked tile. However, quality control based on classification still uses the traditional way by relying on sight. It can increase errors and slow down the process. It can be overcome with artificial visual detectors. It is a result of the rapid development of automation. So to detect defects, this research can use image preprocessing, supervised learning algorithms, and measurement methods. Support Vector Machine (SVM) is used in this study to perform classification, while feature extraction on clay tiles used the Local Binary Pattern (LBP) method. The algorithm is made using python, while for image retrieval, raspberry pi is used. The linear kernel on the SVM algorithm is used in this study. The conclusion in this study obtained 86.95% is the highest accuracy with a linear kernel. It takes 10.625 seconds to classify.

30 June 2021 in Jurnal Rekayasa Sistem & Industri (JRSI)
Verified
Journal Article
Open access
MAMuhammad Nashir Ardiansyah
Muhammad Nashir Ardiansyah
PMPrafajar Sukssesanno Muttaqin
Prafajar Sukssesanno Muttaqin
Murman Dwi PrasetioMurman Dwi Prasetio
Murman Dwi Prasetio
NNNia Novitasari
Nia Novitasari
Aktivitas pemeriksaan persediaan atau Stock-taking merupakan aktivitas pemeriksaan barang manual oleh petugas gudang yang dilakukan secara rutin pada aktivitas pergudangan. Aktivitas ini berfungsi untuk menentukan akurasi persediaan dan mengetahui kondisi persediaan sehingga dapat mengurangi resiko kehilangan, kerusakan, dan keausan persediaan. Aktivitas pemeriksaan persediaan termasuk aktivitas yang memerlukan biaya dan waktu yang besar. Selain itu, aktivitas ini juga tak luput dari kesalahan manusia karena aktivitas pengecekan merupakan aktivitas yang membutuhkan ketelitian tinggi. Penelitian ini bertujuan untuk melakukan identifikasi objek atau produk yang bertujuan untuk menggantikan pemeriksaan manual manusia sehingga proses pemeriksaan jenis dan jumlah barang dapat dilakukan secara otomatis dan presisi. Pengolahan citra digital berbentuk Object Recognition digunakan pada penelitian ini untuk menentukan jenis objek dan jumlah objek. Hasil penelitian menunjukan tingkat deteksi produk tunggal mencapai 90% yang dipengaruhi oleh sudut pengambilan gambar dan tingkat deteksi jumlah objek tunggal mencapai > 81% dengan tingkat pencahayaan yang normal dan sudut pengambilan gambar yang ideal. Diharapkan dengan adanya sistem ini, biaya untuk aktivitas pemeriksaan persediaan dan aktivitas pergudangan secara umum dapat ditekan sehingga efisiensi dan efektivitas dapat dicapai.

11 January 2023 in Buildings
Verified
Journal Article
Open access
PDF file
ZPZhi Yuan Phuah
Zhi Yuan Phuah
Poh Kiat NgPoh Kiat Ng
Poh Kiat Ng
Murman Dwi PrasetioMurman Dwi Prasetio
Murman Dwi Prasetio
KLKia Wai Liew
Kia Wai Liew
BLBoon Kian Lim
Boon Kian Lim
YSYunita Nugrahaini Safrudin
Yunita Nugrahaini Safrudin
Currently, the use of manual labour in the transportation and inspection systems of leading roof tile manufacturing companies in Indonesia is still prevalent. Manual labour is usually labour-intensive, has higher risks of musculoskeletal disorders, and produces frequent occurrences of errors and losses. Furthermore, the current studies of suitable concepts and test protocols for roof tile transportation at the manufacturing stage as well as their inspection systems are not practicable in Indonesia. There is also no study that has used the theory of inventive problem-solving (TRIZ) in the development of concepts and protocols for roof tile transportation and inspection systems. Using TRIZ as a supporting tool, this study investigated the development of a transportation system to be employed during the manufacturing of the roof tile and a test protocol for their usability in Indonesian companies to overcome this concern. The study included screening and scoring concepts and usability test protocols identified from the existing literature, with the support of TRIZ tools such as the engineering contradiction, contradiction matrix, and inventive principles. Thus, the finalised concept comprised a belt conveyor system (Inventive Principle 20: Continuity of Useful Action) with a flipping mechanism for transportation and a vision-based camera for inspection. Results of the study showed that the concept excelled in cost, durability, reliability, versatility, low risk to the product, efficiency, and safety. The t-test protocol (Inventive Principle 23: Feedback) was selected based on the results due to its versatility in testing efficiency, reliability, and productivity. It was concluded that this concept has the potential to alleviate roof tile workers of physical work and reduce the prevalence of musculoskeletal disorders.