GIZMO COP- An Intelligent Security Device for Women Safety with Machine Learning

Abstract

Despite the technological advancements and civilization, issues regarding personal safety especially for women and children never go out of trend. Statistics state that in India, an average of 86 rape happen daily and nearly 49 offences against women are reported for every hour. In such an unsafe society today, it is necessary to carry a smart gadget that ensures safety for women. GIZMO COP has a panic button which upon being activated sends the current location of the victim via SMS with the help of GPS and GSM to the security number fed. In the worst case, when the victim is unable to press the panic button, the temperature and pulse rate of the victim is continuously monitored by sensors and if it changes from the normal pattern or level (when feared) the location details are automatically sent. The threshold values are highly customized using linear regression from machine learning with a set of sensor values related to the user's body condition. In addition to this, the device warns the user from entering into zones that contain hazardous gas and also has sound recording feature for legal evidence. The main advantage of this system is that it operates with cell phone towers, not requiring internet connectivity. With all the wattage values of the individual components, the average battery life is being calculated to approximately 4 hours. Regarding the threshold level for all sensor values, machine learning is implemented to detect the threshold value rather than to use an average value. By this implementation, the accuracy is enhanced by 22% to 28%. With all the basic features required for safety, this device becomes more economical and portable.