IP: 127.0.0.1 On: Fri, 11 Jan 2019 01:56:40 Copyright: American Scientific Publishers Delivered by Ingenta RESEARCH ARTICLE Copyright © 2011 American Scientific Publishers All rights reserved Printed in the United States of America SENSOR LETTERS Vol. 9, 353–358, 2011 Electronic Nose System for Ganoderma Detection A. H. Abdullah ∗ , A. H. Adom, A. Y. Md. Shakaff, M. N. Ahmad, M. A. Saad, E. S. Tan, N. A. Fikri, M. A. Markom, and A. Zakaria Sensor Technology and Application Cluster Research, Universiti Malaysia Perlis, 01000 Kangar, Perlis, Malaysia (Received: 13 November 2009. Accepted: 5 August 2010) An electronic nose (e-nose) is a device that performs electronic olfactory as a possible substitute or to complement human experts. One of its applications is recognition of samples using the detection of the corresponding volatile organic compounds (VOC). There have been continual developments of e-noses due to its growing demand in area such as food quality control, fruit ripeness and plant disease detection. However, the current devices on the market are costly and complicated to operate, as well as having relatively simple algorithms for odour recognition and classification. This paper presents the work of an e-nose development which was designed to be cost effective and uses off-the-shelf components. The e-nose was used to detect the presence of Ganoderma boninense fungus detection from palm oil tree trunks. The prototype utilises an array of metal oxide sensors (MOS) and developed using rapid-prototyping approach. The mobile e-nose offers rapid response, versatility and novelty in the detection of odour using artificial neural network (ANN) algorithm. Initial results show that the el-nose was able to detect the presence of Ganoderma boninense from the tree-trunk odour within a few minutes. This indicates that the system may be used as an early detection system for plant disease monitoring. Keywords: E-Nose, Odour Capturing Module, Microcontroller, PCA, Artificial Neural Network. 1. INTRODUCTION The development of electronic noses (e-nose) as an intel- ligent device that mimics human sensory system has been investigated for a number of years. E-nose is an instrument which comprises an array of electronic chemical sensors with partial specificity and an appropriate pattern recog- nition system, capable of recognising simple or complex odours. 1 The instrument can be used for detection, recog- nition and classification the odour. The e-nose provides a non-destructive approach that can complement the use of human experts in volatile compounds classification. Even though human nose is superior in terms of sensitivity and versatility, the e-nose may be more consistent and can be used to detect harmful or toxic odours. 2 The rapid development of e-nose has been driven by its potential applications and advantages over others approach for odour classification. The classical analysis methods use equipments that are often big, non-portable, as well as requiring complex preparation of the samples and not suit- able for field tests. 3 Most of the applications for e-nose would benefit from a portable system for being easy to ∗ Corresponding author; E-mail: abuhassan@unimap.edu.my use, cost-effective and as flexible as possible: it can eas- ily be trained, programmed and employed for different application. 4 The application of e-nose includes food qual- ity control and assurance, fragrance identification, environ- mental monitoring, and plant disease control. 5 6 Presented is the work conducted on the development of a cost-efficient portable e-nose. The test application for the system looks at the agricultural application. 2. GANODERMA BONINENSE The oil palm (Elaeis guineensis Jacq.) tree is a leading source of cooking oil. Its high demand is also due to the introduction of bio-diesel. In 2008, Malaysia is the world’s second largest palm oil producer, with 17.7 million tonnes of palm oil produced from 4.5 million hectares of land. 7 The oil palm tree is threatened by the Basal Stem Rot (BSR) disease caused by Ganoderma boninense fun- gus as shown in Figure 1, where losses can be significant after repeated planting cycles. BSR has been causing seri- ous damage to the oil palm plantations in Malaysia and is currently causing large amount of losses in revenue. 8 The disease does not show indication in the early stage Sensor Lett. 2011, Vol. 9, No. 1 1546-198X/2011/9/353/006 doi:10.1166/sl.2011.1479 353