Target Seeking Tank Using Unity ML-Agents

Channel:
United States Jasconius Interactive
Subscribers:
553
Published on ● Video Link: https://www.youtube.com/watch?v=0u4tmAvqT-k


Duration: 0:00
168 views
5

Demo of machine learning in Unity using ML-Agents. This tank is trained to seek out the blue target box and drive within 2m of it. This is part of an ongoing development for System Intrusion to use machine learning for enemy AI instead of a hard-coded solution. Next steps are to add ML to shoot and hit the target instead of driving to it, and finally to introduce a second tank for adversarial training.

Written in Unity 2019 and using ML-Agents available at the link below:
https://github.com/Unity-Technologies/ml-agents

"Beauty Flow" Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 4.0 License
http://creativecommons.org/licenses/by/4.0/



Other Videos By Jasconius Interactive


2022-11-24Do more lines of dialog make for a better game?
2022-11-17Dev Log #3 - New Controller and an Old Game
2022-11-10Dev Log #2 - Third Person Multiplayer
2022-11-07Rant: Gotham at 30fps
2022-11-03Dev Log #1 - Catching up
2021-06-13Planet Testing
2021-05-08Procedural Globular Cluster Prototype 2
2021-04-29Procedural Globular Cluster (Prototype)
2020-11-17Target Seeking Tank - Revision 4
2020-10-26Target Seeking Tank - Revision 3
2020-10-16Target Seeking Tank Using Unity ML-Agents
2020-09-23Donut Special - Dev Timelapse
2020-09-07Blueberry Dev Timelapse
2020-09-04HDRP Marble Demo 2
2020-09-04HDRP Marble Demo
2020-07-19Stupid Robots Shooting at Targets
2020-07-10Triple TriJam Dev Timelapse
2020-05-10BouncyBlocks - DOTS Test with Havoc Physics
2020-04-23Virtual Fish Tank - Graphics and Balancing
2020-04-19Ludum Dare 46 - Gameplay and two dead fish
2020-04-19Ludum Dare 46 - Food, Medicine, and Tank Cleaning