Dr. Khaled Hussein

Eng. Haytham Metawie

Ahmed Hany

Ahmed Ehab

Omar Sherif

Omar Saeed

Publishing Date



The term Anosmia stands for the partial or total loss of the sense of smell, and it’s a genuine disease that affects millions of people around the globe. Nowadays, more light has been shed on this ailment because of the surge of the COVID-19 epidemic. This document aims to propose a new system that automatically senses gas leaks, smoke or early fires, spoiled and rotten foods, harmful chemicals, and works as an Air-Quality Control System. Through the aid of an array of gas sensors that each distinguish a certain smell and has unique functionalities. These sensors’ analog pins are connected to the Raspberry-Pi pins which transmit the signals to an ADC module. Furthermore, for processing this data, it shall pass by 2 phases, Pre-Processing phase & Feature Selection, and a Processing Phase to finalize the Data Processing and classify Odors. The System finally notifies the user if any disruptions or any hazardous gas spread occur via a buzzer.

1.1 Purpose of this document

This document is intended to outline the functionality of the E-Nose System, both as a reference for developers and as a quality assurance document for potential clients. The anticipated system is designed to provide effective assistance for Anosmia patients and other consumers in a single device. It shall allow Anosmic people to detect any harmful scents that could lead to undesirable events.

1.2 Scope of this document

This document targets the developing team responsible to construct the device, along with any potential users. They are going to use our complimentary device to monitor multiple gas spreads.

1.3 System Overview

The E-Nose is divided into four sectors. Starting with the hardware components, It is built using Rasp- berry Pi 3. The Raspberry Pi 3 is a micro-controller that is used to gather inputs via various sensors. An array of Mx Sensors is halted on the micro-controller to detect various gases. Depending on the gases detected, the sensor mechanism will generate voltages. The Raspberry Pi will use ADC (Analog to Digital Converter) to convert the signal provided by the sensors from Analog to binary values. Data acquired through the ADC undergoes the pre-processing phase; Which consists of numerous stages. Normalization is the process in which you change the values of numeric columns in the dataset to use a common scale. Then there is the standardization process that involves bringing data into a uniform format that allows people to research the data. The data is then scaled to normalize the range of inde- pendent variables of the data. Then, feature selection is applied by dropping the features we don’t need. After that, we start building our model and processing it, using either Training/Testing Split or Cross- Validation. In the training or testing split, the process entails partitioning a dataset into two subgroups. The training dataset is the initial subset, which is used to fit the model. The second subset is not used to train the model; instead, the dataset’s input element is fed into the model, which then makes predictions and compares them to the estimated parameters. Cross-validation is a resampling strategy for testing and training a model on different iterations that uses different chunks of the data. It’s mainly utilized in situations when the aim is to anticipate how well a predictive model will perform in practice. Furthermore, the Odors are then classified to output the result. The sensors check the output results by comparing them to the data sets. This procedure classifies which gas was detected by the sensors. In the last stage, the signals will be sent to the data set. If any unusual events occur, a notification will be sent to the user.

1.4 System Scope

The project varies between 3 different scopes: 1. Assisting: For Anosmia-Suffering Patients and elderly people by detecting harmful gasses. 2. Smart-Planting: Giving a hand within the Agriculture-field to minimize the percentage error, by Automating the decision time of the fruits and vegetables picking procedure to reduce the percentage error of unripped/over ripped fruits and vegetables. 3. Diagnosing: Aiming to detect several diseases threatening people’s well-being, especially Anosmic people, by classifying dangerous diseases whether they are psychological as ’Schizophrenia’ or physical as ’Azotemia’.