SpaceX's landing simulation employs a reinforcement learning algorithm where a rocket agent optimizes axis control and stability with primary and secondary thrusters to achieve precise landings on a specified pad.
Training Time : 96h
Environment : TraingArea, LandingPads, Agent(Rocket), Buildings, Terrain, Trees
LandingPads : 22
Rockets : 22
Thrusters: (8 balancing), (1 deceleration) each

[Developed & Published] - Vivek Malam - viv3k19 - MARL Simulations
Contact Details : 
Gmail - viv3k.19@gmail.com
GitHub - https://github.com/viv3k19

More information
StatusReleased
PlatformsHTML5
Authorviv3k19
GenreSimulation
Made withUnity
Tagsmarl, mlagents, simulations, spacex, Unity