Abstract

In the industrial world, the speed of the production process plays a very important role in influencing industrial profits. One of the factors that determines production speed is the conveyor. PT. Petrokimia Gresik is one of the industries that uses conveyors, which pays great attention to accuracy and control of material speed. The function of the conveyor is to send raw materials from PT Petrokimia Gresik fertilizer to be processed. Three safety devices are installed on this tool, namely the low speed switch, outlying belt and cable pull switch. However, this complex security system makes it difficult for field technicians to handle conveyor problems. Problems that occur with conveyors, especially in the electrical and instrumentation parts, are sometimes difficult to find. Therefore, the author had the idea to conduct research related to conveyor problem solving analysis using the naive Bayes classification method. The results of the calculation between naive Bayes calculations and expert opinion show 100% agreement, which is an optimal result. It is hoped that these results can help field technicians in diagnosing and resolving problems when problems occur with conveyors.

Analysis o f Damage Handling i n The P T Petrokimia Gresik Factory Conveyor Safety System Using The Naive Bayes Method

Analisa Penanganan Kerusakan Pada Sistem Pengaman Konveyor Pabrik PT Petrokimia Gresik Menggunakan Metode Naive Bayes

Fashoma Yudha Anggana Alkhaqiqi ¹ , Denny Irawan ²

^{1 ,2} Electrical Engineering Study Program ,Muhammadiyah University Gresik , Indonesia

Abstract _ In the industrial world, the speed of the production process plays a very important role in influencing industrial profits. One of the factors that determines production speed is the conveyor. PT. Petrokimia Gresik is one of the industries that uses conveyors, which pays great attention to accuracy and control of material speed. The function of the conveyor is to send raw materials from PT Petrokimia Gresik fertilizer to be processed. Three safety devices are installed on this tool, namely the low speed switch, outlying belt and cable pull switch. However, this complex security system makes it difficult for field technicians to handle conveyor problems. Problems that occur with conveyors, especially in the electrical and instrumentation parts, are sometimes difficult to find. Therefore, the author had the idea to conduct research related to conveyor problem solving analysis using the naive Bayes classification method. The results of the calculation between naive Bayes calculations and expert opinion show 100% agreement, which is an optimal result. It is hoped that these results can help field technicians in diagnosing and resolving problems when problems occur with conveyors.

Keywords: Conveyor; Security System; PT. Petrokimia Gresik; Naive Bayes classification .

Abstrak_Dalam dunia industri kecepatan proses produksi sangat berperan penting dalam mempengaruhi keuntungan industry. Salah satu faktor yang menjadi penentu kecepatan produksi adalah conveyor. PT. Petrokimia Gresik merupakan Salah satu industri yang memanfaatkan conveyor, yang mana sangat memperhatikan akurasi dan kontrol kecepatan material. Fungsi dari conveyor adalah untuk mengirimkan bahan baku dari pupuk PT Petrokimia Gresik yang akan diolah. Pada alat ini terpasang tiga buah perangkat keamanan yaitu low speed switch, outlying belt and cable pull switch. Tetapi dengan sistem pengaman yang kompleks tersebut membuat teknisi lapangan kesulitan saat menangani masalah pada conveyor. Permasalahan yang terjadi pada conveyor khususnya pada bagian electrical dan instrumentasi nya terkad ang sulit untuk ditemukan. Oleh karena itu penulis memiliki ide untuk membuat penelitian terkait analisa troubleshooting conveyor menggunakan metode naive bayes clasification. Hasil dari perbandingan antara perhitungan naive bayes dan pendapat pakar menunjukkan kesesuaian 100% dimana halini merupakan hasil yang opimal. Dengan Hasil ini diharapkan dapat membantu teknisi lapangan dalam mendiagnosa dan mengatasi saat terjadi troubleshooting pada conveyor.

Kata Kunci: Conveyor; Sistem Pengaman; PT. Petrokimia Gresik; Naive bayes classification.

A conveyor belt is a transport medium for a belt conveyor system which is often used to transport goods or materials automatically and efficiently. One of the companies that uses conveyor belts is PT PetroKimia Gresik, which is a company that works in the field of fertilizer manufacturing. However, the conveyor belt at this company often has damage to its safety system. Safety components that are often damaged are the Belt Outlying Sensor, Low Speed ​​Switch Sensor, Cable Pull Switch Sensor, where this damage can have fatal consequences if not treated immediately. Factors causing this damage include dust, vibration, or friction which damages the sensor. Inadequate work environment conditions mean that field workers often find it difficult to determine the cause of damage when the conveyor suddenly does not function properly. This results in disruption of the production process which then impacts all departments [1][2][3].

To make it easier to diagnose the cause of damage and how to treat it, the author plans to use the naive beyes method. The reason the author uses the naive Bayes method is because this method uses the experience of a previous expert to predict future treatment methods. Therefore, field practitioners can trust the results of this calculation and use it as a reference in future work. The following is some research related to conveyors and the naive beyes method that has been carried out previously[4][5].

Research entitled "Problem Solving in Finding Faults (Trouble Shooting) Using the Expert System Method Using Bayesian Theorem on Ship Engines." This research discusses how to solve problems using the expert system method using Bayesian theorem. The Bayesian method calculations are very detailed and can be used as a reference, however the topic that the author raises is not relevant because it discusses ship engines while the author discusses conveyor belts [6].

Previous research entitled "Classification of Impact Damage on a Rubber-Textile Conveyor Belt: A Review" This research discusses the classification of damage that occurs on coal mine conveyors using the Naive Bayes method. This research has been presented well but there is a lack of a table of classification results using the naive Beyes method [7].

From the background above, the author took the initiative to analyze damage to safety sensors caused by dust, environment, vibration and friction on conveyors at PT. Petrokimia Gresik. The aim of carrying out this analysis is to make it easier to handle troubleshooting at the Company.

2.1 Tools

The tools needed to carry out a conveyor belt system analysis are:

1. Digital Multimeter

2. Test Pen

3. Notebook

4. HMI Manual

5. Conveyor System Wiring Diagram Book

6. Manual for all Conveyor Sensors

7. Excel Software

2.2 Method

[ Figure 1 about here.]

The method used for this research is Naïve Bayes, Naïve Bayes Classifier is a classification method that is rooted in classification using probability and statistical methods proposed by the British scientist Thomas Bayes, namely predicting future opportunities based on previous experience so it is known as the Theorem Bayes. The main characteristic of the Naïve Bayes Classifier is the very strong assumption of independence of each condition/event. The Naïve Bayes Classifier method works very well compared to other classifier models. This was proven by Xhemali, Hinde Stone in his journal "Naïve Bayes vs. Decision Trees vs. Decision Trees Neural Networks in the Classification of Training Web Pages” says that “Naïve Bayes Classifier has a better level of accuracy than other classifier models”. On the other hand, with the amount of training data that is not too large, it is quite easy to determine the estimated parameters needed for data classification [8] [9] .

2.3 Solving the Naïve Bayes Method

Solutions that include the Naïve Bayes Classifier method used in expert systems to diagnose conveyor damage in the field. The following are the steps for completing the Naïve Bayes Classifier method

In order to complete the accuracy of the expected results, researchers collected several data on symptoms of conveyor damage which were used as research samples. The following are the results of data collection in the field:

[Table 1 about here.]

Table 1 explains the data on conveyor symptoms that occur in the field. Symptoms This data was obtained from operator complaints related to symptoms of conveyor damage and was given a Symptom code to make identification easier.

[Table 2 about here.]

In Table 2 is data regarding damage that may occur within the scope of conveyor instrumentation and electricity. Damage data in this table is obtained from handling by technicians regarding conveyor damage that has occurred and is given a damage code so that it can be identified more easily.

[Table 3 about here.]

Table 3 explains the data on symptoms based on conveyor damage.

[Table 4 about here.]

Table 4 explains the weight of each conveyor damage that occurs. Damage data in this table is obtained from handling by technicians regarding conveyor damage that has occurred and is given a damage code so that it can be identified more easily.

[Table 5 about here.]

Table 5 explains the possibility of each damage occurring with the specified symptoms.

[Table 6 about here.]

In Table 6, data was obtained by the author from interviews regarding expert recommendations for handling damage to conveyors with the symptoms that have been described.

[Figure 2 about here.]

Based on the data from Figure 2 above, it can be concluded that the accuracy of the calculation results compared with expert opinion is 100%

According to the data presented in Figure 2, precisely in the Bayes Recommendation and Expert Recommendation columns, both of them present the same solution from the 9 trials carried out, which is obtained:

Number of Tests: 9

Number of Matches : 9

Number of Inconsistencies: 0

Percentage: (9/9)*100% =100%

The results above conclude that the percentage of agreement between naïve Bayes calculations and expert opinion is 100%. This is an optimal result so that it can be used as a guideline in overcoming troubleshooting at PT Petrokimia Gresik, specifically on the Factory 2 Production conveyor.

Here I would like to thank those who have supported me in compiling this article:

• Mr. Denny Irawan, S.T., M.T as supervisor
• My parents act as encouragement and motivation
• As well as friends who supported and helped me both materially and in the form of support in the process of preparing this article.

REFERENCES

[1]R. A. Putra, “SIMULASI SISTEM PENGAMAN KONVEYOR BERBASIS PROGRAMMABLE LOGIC CONTROLLER DI PT PLN NUSANTARA POWER PACITAN,” pp. 3–4, 2024.

[2]I. I. Praja, S. S. Dahda, and D. Widyaningrum, “PENERAPAN METODE RELIABILITY CENTERED MAINTENANCE (RCM) PADA PERAWATAN MESIN CONVEYOR UNLOADING PHOSPHATE ROCK (Studi Kasus PT PETROKIMIA GRESIK),” JUSTI (Jurnal Sist. dan Tek. Ind., vol. 1, no. 1, p. 61, 2020, doi: 10.30587/justicb.v1i1.2033.

[3]A. Wisaksono, Y. Purwanti, N. Ariyanti, and M. Masruchin, “Design of Monitoring and Control of Energy Use in Multi-storey Buildings based on IoT,” JEEE-U (Journal Electr. Electron. Eng., vol. 4, no. 2, pp. 128–135, 2020, doi: 10.21070/jeeeu.v4i2.539.

[4]A. Nugroho and Y. Religia, “Analisis Optimasi Algoritma Klasifikasi Naive Bayes menggunakan Genetic Algorithm dan Bagging,” J. RESTI (Rekayasa Sist. dan Teknol. Informasi), vol. 5, no. 3, pp. 504–510, 2021, doi: 10.29207/resti.v5i3.3067.

[5]Rayuwati, Husna Gemasih, and Irma Nizar, “IMPLEMENTASI AlGORITMA NAIVE BAYES UNTUK MEMPREDIKSI TINGKAT PENYEBARAN COVID,” Jural Ris. Rumpun Ilmu Tek., vol. 1, no. 1, pp. 38–46, 2022, doi: 10.55606/jurritek.v1i1.127.

[6]A. Yulianto, K. Dwi Septiady, A. Praja, and S. Mugono, “Pemecahan Masalah Dalam Mencari Kesalahan (Trouble Shooting) Dengan Metode Sistem Pakar (Expert System) Menggunakan Teorema Bayesian Pada Mesin Kapal,” J. Cahaya Bagaskara, vol. 6, no. 1, 2021.

[7]A. R. C. B. A and M. F. Anggamawarti, “Classification of Impact Damage on A Rubber-Textile Conveyor Belt: A Review,” no. 4, pp. 21–27, 2020.

[8]J. Sihombing, “Klasifikasi Data Antroprometri Individu Menggunakan Algoritma Naïve Bayes Classifier,” BIOS J. Teknol. Inf. dan Rekayasa Komput., vol. 2, no. 1, pp. 1–10, 2021, doi: 10.37148/bios.v2i1.15.

[9]Y. Resti, F. Burlian, and I. Yani, “Performance of cans classification system for different conveyor belt speed using naïve bayes,” Sci. Technol. Indones., vol. 5, no. 4, pp. 111–116, 2020, doi: 10.26554/sti.2020.5.4.111-116.

*Corespondent e-mail address Peer reviewed under reponsibility of Muhammadiyah University Sidoarjo, Indonesia

© 2024 Muhammadiyah University Sidoarjo, All right reserved, This is an open access article under the CC BY license()

Received: 2024-04-02

Accepted: 2024-04-16

Published: 2024-04-29

DAFTAR TABEL

Table 1 Tabel ilustrasi Gejala

Table 2 Kode kerusakan dan nama kerusakan

Table 3 Data Gejala Berdasarkan Kerusakan

Table 4 Bobot Kerusakan dan jumlah kemunculan

Table 5 Probabitas kerusakan tiap gejala

Table 6 Data Pengujian Rekomendasi Pakar

DAFTAR GAMBAR

Figure 1 Metode Naive Bayes

Figure 2 Tabel Perbandingan Hasil Pehitungan naïve bayes dan rekomendasi Pakar

References

1. R. A. Putra, “SIMULASI SISTEM PENGAMAN KONVEYOR BERBASIS PROGRAMMABLE LOGIC CONTROLLER DI PT PLN NUSANTARA POWER PACITAN,” pp. 3–4, 2024.
2. I. I. Praja, S. S. Dahda, and D. Widyaningrum, “PENERAPAN METODE RELIABILITY CENTERED MAINTENANCE (RCM) PADA PERAWATAN MESIN CONVEYOR UNLOADING PHOSPHATE ROCK (Studi Kasus PT PETROKIMIA GRESIK),” JUSTI (Jurnal Sist. dan Tek. Ind., vol. 1, no. 1, p. 61, 2020, doi: 10.30587/justicb.v1i1.2033.
3. A. Wisaksono, Y. Purwanti, N. Ariyanti, and M. Masruchin, “Design of Monitoring and Control of Energy Use in Multi-storey Buildings based on IoT,” JEEE-U (Journal Electr. Electron. Eng., vol. 4, no. 2, pp. 128–135, 2020, doi: 10.21070/jeeeu.v4i2.539.
4. A. Nugroho and Y. Religia, “Analisis Optimasi Algoritma Klasifikasi Naive Bayes menggunakan Genetic Algorithm dan Bagging,” J. RESTI (Rekayasa Sist. dan Teknol. Informasi), vol. 5, no. 3, pp. 504–510, 2021, doi: 10.29207/resti.v5i3.3067.
5. Rayuwati, Husna Gemasih, and Irma Nizar, “IMPLEMENTASI AlGORITMA NAIVE BAYES UNTUK MEMPREDIKSI TINGKAT PENYEBARAN COVID,” Jural Ris. Rumpun Ilmu Tek., vol. 1, no. 1, pp. 38–46, 2022, doi: 10.55606/jurritek.v1i1.127.
6. A. Yulianto, K. Dwi Septiady, A. Praja, and S. Mugono, “Pemecahan Masalah Dalam Mencari Kesalahan (Trouble Shooting) Dengan Metode Sistem Pakar (Expert System) Menggunakan Teorema Bayesian Pada Mesin Kapal,” J. Cahaya Bagaskara, vol. 6, no. 1, 2021.
7. A. R. C. B. A and M. F. Anggamawarti, “Classification of Impact Damage on A Rubber-Textile Conveyor Belt: A Review,” no. 4, pp. 21–27, 2020.
8. J. Sihombing, “Klasifikasi Data Antroprometri Individu Menggunakan Algoritma Naïve Bayes Classifier,” BIOS J. Teknol. Inf. dan Rekayasa Komput., vol. 2, no. 1, pp. 1–10, 2021, doi: 10.37148/bios.v2i1.15.
9. Y. Resti, F. Burlian, and I. Yani, “Performance of cans classification system for different conveyor belt speed using naïve bayes,” Sci. Technol. Indones., vol. 5, no. 4, pp. 111–116, 2020, doi: 10.26554/sti.2020.5.4.111-116.