Machine Learning Model to Predict the Severity and Provide an Early Warning of Nerve Agents Threats Using Internet of Things Technologies

Abstract

Nerve Agents (NAs) are often clear, colourless, and tasteless. The only way to identify if a person has been exposed to NAs is through physical symptoms. Longer exposure can lead to death. Rapid identification and alert of NAs involved in any hazardous material incident or terrorist attacks are vital to the protection of first responders as well as for the effective treatment of victims. In this research, high sensitivity NAs sensors that are capable of detecting various types of dangerous chemical agents along with the response measure to the threat will be developed. To ensure a swift and proper response from the authority, this project will focus on Big Data Analytics and IoT streaming data analytics combined with a machine learning model to predict the severity and provide an early warning of the NAs threats using the Internet of Things (IoT) technologies. Built-in IoT sensors in Android devices such as a barometer and other telemetry sensor data will be merged with the NAs sensor data to be sent to a cloud-based database. To ensure rapid, no false positive, and false-negative alerts, Big Data Analytics and IoT Streaming Data Analytics combined with machine learning will be applied to get the most accurate results. This is especially important in any hazardous material incident or terrorist attack so that the most optimal response for the situation can be executed. The complete outcome of this project is in line with Malaysia’s National Defence Policy (MNDP) and will ensure the safety and security of the nation as a whole.