Unity ml-agents used for First Year Google Cardboard VR and PC build Drone Game

For my first year 3D Computer Games Technology assignment, a 3D time trial drone game was produced and developed using machine learning agents in the Unity 2018 game engine and programmed in C# utilising TensorFlow and Python..

The enemy Drone, which fires a raycast laser blast at the users drone, demonstrates the use of pre trained machine learning enemy characters in computer games. The games user interface displays damage to the player from the AI controlled enemy drone. The game can also be played with an X-Box One Controller.

This is probably the first time a Unity 2018 TensorFlow based enemy character, in a 3D computer game compatible with google cardboard and PC, has been submitted as a first year assignment as part of a computer game technology degree course. The aim of using deep learning algorithms in game development is to inspire more students to investigate machine learning modalities.

Below, are screen shots from the submitted game and a link to a compressed file of the finished game for PC.

A team made up of students attending Birmingham City University are planning to develop this game further and produce a 16K VR game that can be played over a 5G network using Edge computing topologies.

One of the challenges we need to overcome is producing an algorithm with the appropriate data structures to efficiently deploy the game for multiplayer network gaming.

The final game will be available for download on the google play store in the 3rd quarter. The beta game file .exe is available in the .7z compression format. The game is currently being developed for Android and PC only and the X-Box One Controller can be used for both formats.



Under the supervision of Prof. Mak Sharma we will be developing a VR game based around the horror for Vodafone 5G. Students form BCU appear in a Vodafone 5G promo with Youtube game streaming legend Tom ‘Syndicate’ Cassell

Super Labrish VR Screenshot and proof of concept with Unity AI ml agents implemented and running inside the game.

Prototype gameplay at runtime.

Project Anwar is officially in development and is based on Whatsapp recordings of a 68 year old man currently living in sunny Jamaica. The laid back monologues chart his journey as a young boy in Birmingham UK after arriving from Jamaica, to rehearsing with Steel Pulse, to his political activism with Lord Anthony Gifford QC to stints in Europe and then back to Jamaica.

Slow playback due to capturing stream and running on low powered laptop. Will review and upload realtime playback soon. This was for the final assessment presentation only. Thank you.

Aston Walker – contact aston.walker@mail.bcu.ac.uk


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Create your website at WordPress.com
Get started
%d bloggers like this: