Embedded E-nose Application to Sense the Food Grain Storage Condition Neha Deshpande and A. D. Shaligram Department of Electronic Science, University of Pune, Pune: 411007, India e-mail: neha.d68@gmail.com , ads@electronics.unipune.ernet,in B. A. Botre, Satish Bindal and S.S . Sadistap AEG, Central Electronics Engineering Research Institute, Council of Scientific and Industrial Research (CSIR), CEERI, Pilani, Rajasthan: 333031, India e-mail: bbotre@gmail.com , sss@ceeri.ernet.in Abstract— With increased demand for food quality and health benefits, need for stringent scrutiny on the inspection of agri- food products have become mandatory. In Indian agriculture, the next challenge is to provide an effective, safe viable storage and handling methods particularly in unpredictable weather conditions. This can be done by using appropriate sensors and system to maintain environmental and storage parameters at predefined level by monitoring of the storage space. The paper presents the application of E-nose system along with smart embedded sensor system to study the deterioration of food grains under different stress (temperature, humidity, insects etc.) and room environmental conditions. The food grain conditions are artificially generated and the effects are studied with Į- Fox 2000 e-nose system. In order to analyze the data of rice, millet, wheat, jawar under different stress conditions, we performed different analysis viz, Principal Component Analysis & Discriminant Factorial Analysis on the acquired E- nose data and results obtained are also presented. Keywords- wireless, food grain warehouse, embedded systems, sensors, electronic nose. I. INTRODUCTION Preservation of the wholesomeness or the quality characters of a food grain commodity is the basic aim of safe storage. The storage of food grains is a complex function of ecological systems comprising of physical, chemical and biological variables. Important variables such as temperature, moisture, oxygen, mites affect grain and its quality. These variables seldom act alone or all at once, rather interact in groups among themselves. When these relationships get disturbed due to unfavorable conditions or defective practices, spoilage occurs. In the present context of sufficient food production in the country, the scientific storage is of paramount importance [1-3]. Temperature and moisture are two main abiotic factors in safe storage of food grains. The effects of these two variables in food grain storage are interrelated. The electronic nose consists of array of metal oxide semiconductor gas sensor, sensor data digitization, feature selection and pattern recognition for the discrimination of samples or items based on odor data [4]. The electronic nose has been used in various other industries such as medicine, environment monitoring, safety & security at airports, space exploration by odor detection and robotics for finding hazardous chemicals. Electronic noses are finding wide range of applications in analyzing food items, their quality inspection and processing monitoring [5-8]. Considering the wide potential applications of electronic nose tool, the paper presents the application of E-nose system along with smart embedded sensor system to study the deterioration of food grains under different stress (temperature, humidity, insects etc.) and room environmental conditions. The food grain conditions are artificially generated and the effects are studied with e-nose system present at Agri- Electronics Group, CEERI Pilani. In existing sensing chamber of E-nose module available, it is not possible to directly monitor temperature and humidity of the grain under study. We modified the sensing chamber lid of E-nose system so as to monitor temperature as well as humidity of different grains under different conditions along with the head space data. Smart embedded sensor system along with wireless connectivity developed by AEG, CEERI is used to monitor room temperature gradient and humidity conditions. Further in order to analyze the data of rice, millet, wheat, jawar, red gram dal under different stress conditions, we performed different analysis viz. Principal Component Analysis & Discriminant Factorial Analysis on the acquired E-nose data. II. SYSTEM DESCRIPTION In our work we have used the e-nose ‘Į- Fox 2000’. Different grains such as rice, wheat, red gram dal, great millet and bulrush millet were used for testing at various conditions such as different temperatures, moisture content, pest infected, with foreign material and in simple glass container as well as in ‘bori’ (sac). Humidity sensor SY HH 220 was used to measure humidity of grain and temperature of grain is measured using sensor LM35. PIC microcontroller embedded system board is used to interface these sensors with PC [9-11]. When the e-nose is set in the Gas mode, no internal carrier gas flows through the injector and over the solid sorbent trap during the sample time. The internal pump draws gas samples into the instrument through the sample inlet and into the trap. There are various sensors in the system that sense different components of gases in the head space generated above the grain sample. This data is further analyzed by the e-nose. Thus original clean dry fresh grain sample data is recorded and then the variations in the grain conditions are recorded. The Principal Component Analysis (PCA) and Discriminant Function Analysis (DFA) clearly show the differentiation in these states. While working on this system 2010 International Conference on Computational Intelligence and Communication Systems 978-0-7695-4254-6/10 $26.00 © 2010 IEEE DOI 10.1109/CICN.2010.120 608 2010 International Conference on Computational Intelligence and Communication Networks 978-0-7695-4254-6/10 $26.00 © 2010 IEEE DOI 10.1109/CICN.2010.120 608