Smart Security System using ML

To build a safety device for women

Created on 20th September 2020

•

The project aims to provide low cost IoT based solutions for women safety which includes :--->

Electric shock gadget for defence to help women escape in critical situation.
In case of voice recognition system fails the device can also be turned on manually by pressing a switch

Advantages of Gadget over existing Technology :--->

Portable Can be easily carried any where.
l do the predictive analysis of the past 10 years crime and thus detect the crime hotspot and also live map with hotspot and the results are displayed in
In case the device gets broken the vibration sensor sends the last location to the registered phone number.

Hurdles we can into while building this project :--->

It was very difficult to secure the information of our ML model result as we wanted to make sure except police no one could access it.
Another bug we faced if a person by mistake turns on the device or make a voice command the device gets activated automatically which would in turn create unnecessary chaos.

How did we overcome it ?

We secured our ml model by creating a login window with no registration button in it and only the head of the police department can decide who can access it by giving detail in user registration excel file.
We used a vibration motor which vibrates for a window span of 5 second to let the person the know the device is activated and the person can turn off the device by pressing on terminal button if device gets turned on by mistake preventing from creating unnecessary chaos.
The image stored in sd card can be sent to police department through g-mail with the help of portable wifi.

Technologies used

Machine Learning

Internet of Things (IoT)