Application of Smart Inspection with Statistical Process Control in Visual Quality Inspection of Tempeh

Aisyah, Aisyah (2026) Application of Smart Inspection with Statistical Process Control in Visual Quality Inspection of Tempeh. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5010211129-Undergraduate_Thesis.pdf Download (783kB)

Abstract

Tempe, sebagai makanan tradisional Indonesia berbahan dasar kedelai fermentasi, merupakan produk yang banyak dikonsumsi serta memiliki nilai gizi dan budaya yang tinggi. Namun demikian, proses inspeksi kualitas pada usaha kecil dan menengah (UKM) masih dilakukan secara manual, sehingga menimbulkan ketidak konsistenan, subjektivitas, dan inefisiensi. Meskipun inspeksi kualitas memiliki peran penting, penelitian mengenai pengembangan kerangka inspeksi visual berbasis kecerdasan buatan (AI) yang terjangkau dan sesuai dengan kebutuhan UKM pangan tradisional masih sangat terbatas. Penelitian ini bertujuan untuk mengisi kesenjangan tersebut dengan mengusulkan sistem inspeksi visual berbantuan AI berbiaya rendah untuk tempe, yang mampu mendeteksi cacat berdasarkan tekstur, distribusi granula, warna, dan cakupan miselium. Metodologi penelitian meliputi pengambilan citra secara terkontrol, optimasi perintah (prompt) AI, serta penerapan alat Statistical Process Control (SPC) seperti Xbar-S Chart, NP-Chart, dan C-Chart untuk mengevaluasi dan memantau kualitas produk. Data diperoleh dari foto terdokumentasi sendiri dengan 40 kali replikasi guna memastikan reliabilitas statistik. Pengambilan citra terkontrol dan perintah AI yang terstandarisasi digunakan untuk mengklasifikasikan cacat berdasarkan tekstur, sebaran granula, warna, dan cakupan miselium. Diagram SPC (Xbar S-Chart, NP-Chart, dan C-Chart) digunakan untuk memantau stabilitas proses, sementara uji ANOVA mengonfirmasi bahwa variasi dalam deteksi cacat signifikan secara statistik, sehingga memvalidasi responsivitas sistem AI. Dengan mengoptimalkan pengaturan pencahayaan, khususnya menggunakan lampu kilat (flashlight) smartphone yang diposisikan secara sentral di atas sampel, bayangan dapat diminimalkan sehingga deteksi menjadi konsisten. Secara keseluruhan, inspeksi berbantuan AI menawarkan bagi UKM solusi yang andal, terjangkau, dan dapat diskalakan untuk meningkatkan kualitas produk, meminimalkan kerugian finansial, serta memperkuat daya saing dalam industri tempe.
====================================================================================================================================
Tempeh, a traditional Indonesian fermented soybean food, is widely consumed and valued for its nutritional and cultural significance. However, quality inspection in small and medium enterprises (SMEs) is still largely performed manually, leading to inconsistency, subjectivity, and inefficiency. Despite the importance of quality inspection, there has been little research on developing affordable, AI based visual inspection frameworks tailored for traditional food SMEs. This study aims to address that gap by proposing a low cost AI assisted visual inspection system for Tempe, capable of detecting defects based on texture, granule distribution, color, and mycelium coverage. The methodology involves controlled image capture, AI prompt optimization, and the application of statistical process control (SPC) tools such as Xbar S Chart, NP Chart, and C Chart to evaluate and monitor product quality. Data are collected from self documented photos with 40 replications to ensure statistical reliability. Controlled image capture and standardized AI prompts were used to classify defects based on texture, granule dispersion, color, and mycelium coverage. SPC charts (Xbar-S Chart, NP-Chart and C-Chart) monitored process stability, while ANOVA confirmed that variations in defect detection were statistically significant, validating the responsiveness of the AI system. By optimizing lighting arrangements, specifically using a smartphone flashlight positioned centrally above the sample, shadows were minimized, enabling consistent detection. Overall, AI assisted inspection offers SMEs a reliable, affordable, and scalable solution to enhance product quality, minimize financial losses, and strengthen competitiveness in the tempeh industry.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Artificial intelligence, quality control, defect detection, statistical process control, low-cost inspection. Artificial Intelligence(AI), pengendalian kualitas, deteksi cacat, produk, pengendalian proses statistik, inspeksi berbiaya rendah.
Subjects:	H Social Sciences > HD Industries. Land use. Labor > HD3656 Inspection. Factory inspection T Technology > TS Manufactures > TS156 Quality Control. QFD. Taguchi methods (Quality control)
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Industrial Engineering > 26201-(S1) Undergraduate Thesis
Depositing User:	Aisyah Aisyah
Date Deposited:	31 Jul 2026 01:42
Last Modified:	31 Jul 2026 01:42
URI:	http://repository.its.ac.id/id/eprint/138316

Actions (login required)

View Item