Design of an IoT-Based Smoke Detector For Early Warning of Lithium-Ion 18650 Battery Explosion in Electric Vehicles

Abstract

Lithium-ion batteries are widely used as the primary energy source in electric vehicles due to their high energy density and long service life; however, they present a risk of thermal runaway that can result in smoke, fire, or explosion when the battery management system fails to interrupt an overcharging process. This study aims to design and evaluate an Internet of Things (IoT)-based smoke detector prototype for providing early warning of lithium-ion 18650 battery failure caused by overcharging. An experimental method was employed using a prototype equipped with an MQ-7 gas sensor to detect carbon monoxide (CO) concentration, an Arduino Uno microcontroller for data processing, and a NodeMCU ESP8266 module for transmitting real-time notifications through the Blynk application, supported by an LCD display, LED indicators, and a buzzer alarm. The experiment was conducted by charging 18650 lithium-ion battery cells with a constant current of 3 A until an explosion occurred, using battery samples with initial states of charge (SOC) of 0% and 50%. The results indicate that the MQ-7 sensor successfully detected CO gas released during battery explosions, with concentrations ranging from 27.14 ppm to 131.55 ppm, while calibration against a CO meter showed an average error of 2.25%. Batteries with a 0% SOC produced higher CO concentrations than those with a 50% SOC due to a longer overcharging duration before explosion. The study concludes that the developed prototype can function effectively as an early warning system to reduce fire risks associated with lithium-ion battery thermal runaway in electric vehicles.