HackOverflow 2024 Project

Peekabot is an innovative, cloud-connected robot that revolutionizes home safety. Built on Arduino and Raspberry Pi platforms, it utilizes advanced motion tracking technology to monitor your home environment. Peekabot provides real-time alerts for various scenarios, from child safety to elderly care and home security.

• Advanced Motion Tracking: Utilizes state-of-the-art technology to track movement of people and objects in real-time.
• Instant Alerts: Sends immediate text notifications for critical situations (e.g., toddler in danger, elderly falls, potential intruders).
• Multi-Purpose Design: Adaptable for child monitoring, elderly care, and home security.
• Remote Access: Cloud connectivity allows you to check on your home from anywhere.

• Hardware:

  • Arduino Uno R3
  • Raspberry Pi 4
  • 4 motors for mobility
  • Circuit board for integration

• Software:

  • OpenCV and Mediapipe
  • Local web server for remote access
  • Cloud-based alert system via AWS IoT

The project is divided into multiple components to manage different functionalities and hardware integrations:

Contains the code for the two Arduino boards that control the left and right wheels of Peekabot.

• Arduino_1.ino: Code to control the left wheels of the robot. This script receives commands from the Raspberry Pi to move the left motors accordingly.
• Arduino_2.ino: Code to control the right wheels of the robot. Similar to Arduino 1, this script moves the right wheels based on commands received.
• README.md: Documentation for setting up and using the Arduino boards in the Peekabot project.

Contains documentation and configuration information for setting up AWS IoT and ngrok for the Peekabot project.

• README.md: Details how to configure AWS IoT Core for secure command and communication handling. It also includes steps to set up ngrok for live video streaming accessible through the subdomain peekabot.uverma.com.

This folder includes the computer vision code that allows the robot to detect and track human movements.

• distance_estimator.py: A utility script used to calculate the distance of detected individuals from the camera. It helps determine whether the robot should move forward, backward, or stay.
• main.py: A general-purpose script that can be used on any machine with a camera (e.g., a laptop or desktop) to run the pose detection and generate movement commands. It displays both bit representations and text descriptions of the detected commands for easy debugging.
• pi_main.py: Specifically designed to run on the Raspberry Pi, this script connects to AWS IoT to receive commands and uses the camera feed to control the robot's movement in real-time. It communicates with the Arduinos to execute the movement instructions.

1. Environment Scanning: Continuously monitors the designated area using advanced sensors.
2. Data Processing: Analyzes collected data in real-time using Mediapipe to identify potential risks or unusual activities.
3. Alert Generation: Instantly creates and sends text alerts via AWS IoT when predefined safety thresholds are crossed.
4. Remote Monitoring: Users can check in and control Peekabot from anywhere via a secure web interface, accessible through ngrok at peekabot.uverma.com.

Peekabot utilizes AWS IoT Core to handle communication between the Raspberry Pi, Arduinos, and other devices. This integration allows for:

• Remote Control: Sending commands securely from the cloud to the robot.
• Cloud-Based Alerts: Instant alert generation when the robot detects unusual activity.

The ngrok live feed provides a real-time video stream of the robot's camera feed, which can be accessed through the subdomain peekabot.uverma.com. This feature allows users to monitor the robot's environment and activities from anywhere.

• Create an IoT Thing: Set up your device in the AWS IoT Core dashboard.
• Attach Certificates: Download and attach the necessary certificates to ensure secure communication.
• Configure Endpoint: Use the AWS endpoint in pi_main.py to connect your Raspberry Pi to the AWS IoT platform.

• Install ngrok on the Raspberry Pi.

• Run ngrok with the command:

  ngrok http 5000

• The live stream is hosted at peekabot.uverma.com for remote viewing of the camera feed.

• AI-Powered Predictive Alerts: Use AI algorithms to predict and notify users of potential risks before they occur.
• Smart Home Integration: Seamlessly connect with other IoT devices for a comprehensive home automation experience.
• Customizable Behavior Patterns: Allow users to define specific monitoring rules tailored to their unique needs.

Experience Peekabot in action! Scan the QR code below to access our live demo from your smartphone. Receive real-time updates and interact with Peekabot's interface to explore its capabilities.

This project is open-source and available under the GNU License.

Feel free to contribute to this project by opening an issue or submitting a pull request.

Peekabot: Keeping an eye on what matters most, so you don't have to.