Skip to content

HondaPL/Inteligent-Garbage-Truck

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Inteligent Garbage Truck AI

Table of contents

Task Description

The task of an inteligent garbage truck is to autonomously gather waste from dustbins. The truck should recognize the types of dustbins, look inside the dustbins and check if the garbage is correctly sorted, empty the bins, move the waste to the landfill and offload it in a suitable place.

Our solution

To solve this task we used:

  • pygame for agent environment representation
  • BFS algorith to implement uninformed search strategies in our project
  • A* algorith to implement informed search strategies in our project
  • Learning Methods:
    • Decision Trees (we used them to make decision if the garbage truck should take the trash or not by specific arguments like:
      • Trash size
      • Trash weight
      • Space on the trash yard
      • And much more
    • Neural Networks (we used them to recognize type of trash by its image)
    • Genetic Algorithms (we used it to solve Travelling salesman problem)

Screenshots

Example screenshot Example screenshot Example screenshot

Technologies

  • python
  • pygame library (For environment representation)
  • sklearn library (For Decision Trees)
  • keras library (For Neural Networks)

Setup

  1. Run main.py
  2. Press space to start the project :)
  3. If project didn't load you need to import some libraries: like pytorch, pygame, sklearn, keras and some more. To do it a recommend to use pip eg. pip install torch or pip3 install torch.

Features

  • Move garbage truck with arrow (up to move forward, left and right to rotate the agent).
  • Go to specific square on map by pressing left mouse button and then right mouse button.
  • Toggle algorith from A* to BFS by pressing A.
  • Place obstacles on map by selecting square by left mouse button and then pressing O.
  • Place trash bins on map by selecting square by left mouse button and then pressing T.
  • Run whole app by pressing space.
  • Trash images recognition created with neural networks.
  • TSP problem solved with genetic algorith.
  • Make decission whether to pick up trash or not by many arguments.
  • Nice HUD to show all important informations to user.
  • Show trash image when collecting.

Status

Project is: completed

Contact

Created by :

2021