DEVELOPMENT OF ELECTRONIC NOSE PROTOTYPE FOR BANANA RIPENESS MEASUREMENT

• Main Author: NUR FAHMAN (NIM : 13313055)- DIMAS INDRA CAHAYA RAMADHAN (NIM : 13313081), KHALID
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/22814
• Institution: Institut Teknologi Bandung

Determination of banana maturity is important in determining harvest time, distribution time, and storage life. Peel color inspection method using human vision has disadvantage. Peel color and fruit flesh is not always correlated. For example, banana which is dipped in additive coating has yellow color that is not correlated with fruit flesh maturity. Thus, there is a need to develop instrument that can measure and classify banana ripeness using other parameter. In this study, electronic nose prototype was developed for banana maturity classification. Electronic nose consist of sampling system, array of gas sensor, and pattern recognition algorithm. Headspace gas sampling was used in this prototype. VOC-1 (ethylene), VOC-2 (alcohol), and VOC-3 (Benzene, ammonium, CO2), were used in this prototype. Prior to measurement, ethylene sensor was calibrated using Ethylene Spy ES-100 and percentage error of 6.02 ± 2.14 % was obtained. After calibration, measurement was conducted to 25 samples of Ambon Bananas. Sensor response in form of steady state voltage was then identified and processed using Principal Component Analysis and Artificial Neural Network. Loading plot shows that has sensor optimal number were two, which were combination of VOC-1 and VOC-2, or VOC-1 and VOC-3. Electronic nose was sucessfully classified 23 bananas (92%) into approtiate ripeness stage.