question

In general, we want people to be able to connect FlexSim to any RL platform they want, including SageMaker. Based on this documentation, it looks like it is already possible to connect FlexSim and SageMaker, since SageMaker appears to use OpenAI Gym.

SageMaker documentation about using a simulator in RL: https://docs.aws.amazon.com/sagemaker/latest/dg/sagemaker-rl-environments.html

FlexSim documentation on creating an OpenAI Gym simulator for RL: https://docs.flexsim.com/en/23.0/ModelLogic/ReinforcementLearning/WorkingWithGym/GettingStarted/GettingStarted.html

My guess is that you'll need to learn how to make a Docker container with a FlexSim model being run as a Gym environment. That can be tricky, but it should be possible. In version 22.2, users had to create custom Docker containers running a certain python script. While the documentation is aimed at Bonsai, you can likely learn what you need from that documentation:

https://docs.flexsim.com/en/22.2/ModelLogic/ReinforcementLearning/WorkingWithBonsai/GettingStarted/GettingStarted.html

We added integration with Bonsai because many independent customers asked for it. If another platform gets as much interest as Bonsai, we would investigate adding support for that platform as well.