Dry Cannabis Detection by Using Portable Electronic Nose

The cannabis, also known as marijuana, is one of the psychoactive drugs from the cannabis plant. Dry cannabis is mostly found in Sumatera, Indonesia. The cannabis has psychoactive properties which are cannabis saliva, cannabis indica, and cannabis ruderalis. Dry Cannabis is usually hide by drug smuggler inside unpredicted location during delivery. It is difficult to find if only use visual inspection. The visual inspection is usually combined with the X-Ray in airport. The common method to identify the dry cannabis is using the Trained K-9 dogs. It is important to has the portable dry cannabis detection when agent work on field. This research has aim to mimic the K-9 dogs' ability to smell the dry cannabis aroma by deploying the electronic nose (e-Nose). The electronic nose (e-Nose) technology is able to learn and identify aroma or odor based on the trained model. The model is created by the training data is created from recorded aroma sample of dry cannabis. The e-Nose consist of sensor array, embedded device and artificial intelligence. This research is initial research creating a Smart Cannabis Detection System. The proposed system combined an e-Nose with Internet of Things which consist of end-device and server. The proposed end-device is designed as a portable device. The end-device is formed by Arduino, ESP-01, and sensor array. The server is made of jetson nano. The server works locally through internet connection. The dataset is created by using 3000 of dry cannabis aroma. The accuracy of the dry cannabis prediction is 0.96 by using artificial neural network (ANN).