Pengendalian Kualitas Produksi Portland Composite Cement (PCC) di PT XYZ Menggunakan Diagram Kendali Simultan Max-Half-Mchart Berbasis Minimum Regularized Covariance Determinant (MRCD)

Fidela, Sabina Aisha (2025) Pengendalian Kualitas Produksi Portland Composite Cement (PCC) di PT XYZ Menggunakan Diagram Kendali Simultan Max-Half-Mchart Berbasis Minimum Regularized Covariance Determinant (MRCD). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003211010-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (1MB) | Request a copy

Abstract

Pengendalian kualitas adalah elemen penting dalam industri manufaktur untuk menjaga konsistensi produk dan memenuhi standar mutu yang ditetapkan. Kualitas semen yang baik terutama pada Portland Composite Cement (PCC) sangat krusial dalam menentukan kekuatan, daya tahan, dan keandalan struktur bangunan. Penelitian ini bertujuan untuk melakukan monitoring kualitas produksi semen PCC di PT XYZ dengan menggunakan metode Max-Half-Mchart berbasis Minimum Regularized Covariance Determinant (MRCD). Diagram kendali ini memiliki kemampuan untuk memantau berbagai karakteristik kualitas secara simultan sehingga dapat memberikan gambaran menyeluruh mengenai stabilitas proses produksi. Metode Max-Half-Mchart dirancang untuk mendeteksi variasi yang tidak teridentifikasi oleh diagram kendali konvensional, terutama pada data multivariat. Penelitian ini berbasis Minimum Regularized Covariance Determinant (MRCD) yang dirancang untuk mengestimasi kovarians yang lebih stabil pada data multivariat, terutama pada dengan jumlah karakteristik kualitas yang cenderung besar atau data yang mengandung outlier. Pengendalian kualitas semen PCC di PT XYZ menggunakan diagram kendali Max-Half-Mchart pada data Fase I & II ditemukan adanya pengamatan out of control yang disebabkan oleh pergeseran rata-rata, variabilitas, serta variabilitas & rata-rata proses secara simultan. Setelah dilakukan analisis lebih lanjut, terlihat karakteristik kualitas yang paling banyak menyebabkan data out of control adalah Blaine yang disebabkan oleh variansi konsentrasi yang cenderung tinggi. Selain itu, dilakukan perbandingan metode Max-Half-Mchart konvensional dengan Max-Half-Mchart berbasis MRCD, Det-MCD, dan Fast-MCD. Hasil perbandingan keempat metode tersebut menunjukkan bahwa Max-Half-Mchart berbasis MRCD lebih sensitif dalam mendeteksi pergeseran terutama pergeseran variabilitas & rata-rata proses secara simultan. Hasil perhitungan indeks kapabilitas secara multivariat diperoleh nilai MPp sebesar 0,97 dan sebesar MPpk 1,65 yang dapat disimpulkan bahwa secara keseluruhan rata-rata proses sudah mendekati target spesifikasi dan telah memenuhi spesifikasi yang telah ditetapkan. Namun nilai MPp yang masih dibawah 1,00 mengindikasikan variabilitas antar data masih cukup besar.
========================================================================================================================================
Quality control is a critical aspect of the manufacturing industry to maintain product consistency and meet established quality standards. The quality of Portland Composite Cement (PCC) is particularly crucial, as it directly affects the strength, durability, and reliability of construction structures. This study aims to monitor the production quality of PCC at PT XYZ using the Simultaneous Max-Half M-Chart method based on the Minimum Regularized Covariance Determinant (MRCD). This multivariate control chart enables the simultaneous monitoring of multiple quality parameters, offering a comprehensive overview of process stability. The Max-Half M-Chart is designed to detect process shifts in mean and variability that may not be identified by conventional charts, especially in the presence of outliers. Applied to Phase I and II data, the control chart identified several out-of-control observations resulting from shifts in mean, variability, and both simultaneously. Further analysis revealed that the most frequent contributors to process instability is Blaine largely due to high variability in concentration. A comparative analysis of the conventional Max-Half M-Chart with its robust variants—MRCD, Det-MCD, and Fast-MCD—showed that the MRCD-based version was the most sensitive in detecting simultaneous shifts. The multivariate process capability indices obtained were MPp = 1,88 and MPpk = 1,23 which can be concluded that the overall average process is close to the target specification and has met the predetermined specifications. However, the value of MPp which is still below 1.00 indicates that the variability between data is still quite large.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Pengendalian Kualitas, Diagram Kendali Simultan, Max-Half-Mchart, Industri Semen, Minimum Regularized Covariance Determinant, Quality Control, Simultaneous Control Chart, Max-Half-Mchart, Cement Industry, Minimum Regularized Covariance Determinant
Subjects:	T Technology > TS Manufactures > TS156 Quality Control. QFD. Taguchi methods (Quality control)
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Sabina Aisha Fidela
Date Deposited:	01 Aug 2025 01:48
Last Modified:	01 Aug 2025 01:48
URI:	http://repository.its.ac.id/id/eprint/125483

Actions (login required)

View Item