Skip to content

Latest commit

 

History

History
80 lines (51 loc) · 3.17 KB

README.md

File metadata and controls

80 lines (51 loc) · 3.17 KB

pestOVERFLOW

An IOT pest detector which aims to do pest analysis around your home pestoverflow

link to the video and devpost

https://devpost.com/software/ap-pest-overflow

Slides Link

Inspiration

  • inspired from stack overflow(a website for to help you debug your programming problems)
  • we came out with the idea pest overflow(an IOT device which help to debug your pest problems at home)

Logo Inspiration

  • We are so good at detecting the pest that the roaches overflow from the dustbin.

Marketing statement

with rising humidity and global warming more and more pest is coming to your home, how do you sleep at night knowing that there are scary creatures lurking around your home. Fear not, pest overflow is like your real life ghostbusters that aims to detect the level of pest u have at your home.

  • you still have to call your own pest exterminator after that.

What it does

Pest Overflow(product name) is a realtime pest detection tool.

(1) It helps to detect pest using a motion sensor

Source Code from Twilio source code twilio api

(2) once detected, it will take an image of the pest, showing its last location

Autentication Successful on Raspberry Pi successful send

(3) An sms is being sent out to the user to update there is a pest at home

Mobile Phone Home Notification mobilephone_1_edited

Messaged Received on Iphone mobilephone2_edited

    • user can take action to exterminate the pest or call for pest exterminator after that.**

How I built it

Hardware

  • we used the raspberry pi 3, where we house all the IOT sensors on it
  • we used the raspberry pi camera, to take an image of the last location of the pest
  • we used the PIR Sensor(motion sensor), to detect the pest moving around the house
  • we used an Led Sensor, to have a feedback mechanism that the PIR sensor is working

Software

  • We used python 3 to write the logic for the pest detector
  • We used twilio API for update the user, when there is a pest

Challenges I ran into

  • We had the challenge of making the machine learning pest detection work on the raspberry pi.
  • we had the challenge of using the twilio API because it was free, we can only send text messages for the trial version and unable to send an MMS of the image.

Accomplishments that I'm proud of & what i have learnt

  • It is our first time working with the raspberry pi and using the sensors in the hackathon.
  • It is also our first experience building and connecting the API to the raspberry pi.

What's next for AP - Pest Overflow

To build a robo exterminator, to catch the bugs