Anti-viRL

Covid-AR-Shooting-RL (Anti viRL)

Github Repository

This project won the third prize at the Covid-19 Gameathon organized by Angelhack.

Certificate offered by the organizers

Introduction

Problem Statement

In an attempt to fight the virus, this hackathon calls upon developers to create innovative games that address issues faced by people during the pandemic. We are supposed to design a game that will help bring about awareness about the on going pandemic.

The Solution

Our game takes the essentials of hygiene practices such as frequent sanitizing, social distancing, protecting oneself with a mask etc. into a fun and exciting PC and mobile cross-platform FPS, surviving with the help of others (yes, it is a multiplayer through and through, without the support for others, the game will end a little too quick) and fending against the swarms of COVID virus cells and each other.

Team

  • Adhiraj Bhagawati
    •
  • Adhyan Mall
    •
  • Tamizhiniyan S R
    •
  • Rohit Ramesh
    •
  • Pallavi Y
    •

My Role

I was responsible for the Multiplayer aspect with Photon (PUN 2), and making it work with Google ARCore. I was also responsible for implementing all the game mechanics.

Tech Stack

  • Unity3D
  • Google AR Core & AR Foundation
  • PhotonSDK
  • Unity ML Agents
  • Affinity Designer
  • Blender
The RL training using ML Agents

Modules

We planned the solution in a modular fashion, the key modules being:

  • Augmented Reality Environment
  • The Networking Module
  • The PC gameplay module
  • Cross-platforming
  • Reinforced Learning in the enemies (Covid-19 particles)

Game Design

  • The player is equipped with a sanitizer gun which is used to shoot the virus particles.
  • If the player comes in contact with a particle, it will respawn.
  • Going too close to the particle reduces players health.
  • Shooting other players with sanitizer results in the increase of health but approaching another player results in the violation of social distance.
  • Masks are present and collecting them increases health of the player.
  • The Reinforced Learned COVID particles have a win/lose reward state which makes the floor red in the player infected state and green if the player successfully evades the particles.
  • As a survivalistic aspect, only the other players can view your player’s health and vice versa that enables them to help each other to fend off against the onslaught of COVID virus cells.