Electronic Theses and Dissertation

Near infrared reflectance spectroscopy (NIRS) merupakan salah satu metode non destruktif yang paling menarik dan biasa digunakan untuk analisis di berbagai bidang, termasuk bidang veteriner. Pengujian kadar lemak pada daging sapi dapat dilakukan di laboratorium menggunakan campuran bahan kimia, membutuhkan waktu yang lama sehingga tidak efisien, dan membutuhkan biaya yang relatif mahal. Penggunaan near infrared reflectance spectroscopy (NIRS) merupakan teknologi alternatif yang dapat menggantikan pengujian di laboratorium yang dapat menganalisa dan memberikan informasi mengenai kadar lemak pada daging sapi. Penelitian ini bertujuan untuk menghasilkan spektrum pendugaan kadar lemak pada daging yang diawetkan dengan kemampuan teknologi near infrared reflectance spectroscopy (NIRS). Perangkat NIRS akan didapatkan raw spectrum yang telah diakuisisi kemudian spektrum dianalisis data menggunakan partial least square regression (PLSR). Sampel yang digunakan adalah sebanyak 250 g daging sapi aceh bagian paha belakang yang dibagi menjadi 30 bagian dengan setiap bagian memiliki berat 8 g. Penelitian dilakukan menggunakan Rancangan Acak Lengkap (RAL), yang terdiri dari 3 perlakuan dimana daging sapi aceh direndam dalam larutan cuka aren 0%, 2,5%, dan 5% selama 24 jam pada suhu refrigerator. Setiap perlakuan dilakukan pengulangan sebanyak 10 kali. Berdasarkan hasil penelitian, analisis data dengan metode partial least square regression (PLSR) memungkinkan NIRS untuk menentukan komponen target dalam sampel berdasarkan spektrum NIRS yaitu kadar lemak dengan memperoleh nilai koefisien korelasi (r) yang didapat yaitu 0,63, koefisien determinasi (R²) yaitu 0,409 dan RMSEC yaitu 0,310. Kesimpulan dari penelitian ini adalah bahwa metode NIRS dapat digunakan untuk mendeteksi kadar lemak daging sapi yang diawetkan, hal ini dibuktikan dari nilai koefisien korelasi, koefisien determinasi, dan RMSEC kadar lemak yaitu (0,63; 0,409; dan 0,310) tergolong pada prediksi yang masih kasar (sufficient permormance).

Kata Kunci : Cuka Aren, Daging Sapi, Kadar Lemak, Near Infrared Reflectance Spectroscopy, Partial Least Square Regression

Near infrared reflectance spectroscopy (NIRS) is one of the most interesting non-destructive methods and is commonly used for analysis in various fields, including the veterinary field. Testing fat content in beef can be done in a laboratory using a mixture of chemicals, takes a long time so is inefficient, and relatively expensive. The use of near infrared reflectance spectroscopy (NIRS) is an alternative technology that can replace laboratory testing and can analyze and provide information about the fat content in beef. The aim of the research was to produce a spectrum for estimating fat content in cured meat using near infrared reflectance spectroscopy (NIRS) technology. The NIRS device obtains the raw spectrum that has been acquired, then the spectrum data will be analyzed using partial least squares regression (PLSR). The sample used was 250 g of Acehnese beef from the back of the thigh which was divided into 30 parts with each part weighing 8 g. The research was carried out using a Completely Randomized Design (CRD), which consisted of 3 treatments where Acehnese beef was soaked in a solution of 0%, 2.5%, and 5% palm vinegar for 24 hours at refrigerator temperature. Each treatment was repeated 10 times. Based on the research results, data analysis using the partial least squares regression (PLSR) method allow NIRS to determine the target components in the sample based on the NIRS spectrum, namely fat content by obtaining the correlation coefficient (r) values obtained, namely 0.63, coefficient of determination (R²) is 0,409 and RMSEC is 0.310. The conclusion of this research was that the NIRS method can be used to detect the fat content of cured beef, this is proven by the correlation coefficient, coefficient of determination, and RMSEC fat content values, namely (0.63; 0.409; and 0.310) which are classified as good predictions still rough (sufficient performance). Keywords: Beef, Fat Content, Near Infrared Reflectance Spectroscopy, Palm Vinegar, Partial Least Square Regression.