Current Status of rl-testbed-for-energyplus

[FrancisBases]

Hello, i'm an experienced machine learning developer, and i'm considering using rl-testbed-for-energyplus for the framework i am designing for urban simulation in the scope of my phd thesis.

Goal: Show or compare the benefits of RL related to the classic control system approach for controlling actuators (lights and venetian blinds) with rewards from sensor states(classroom sensors) using the simplest possible example.

We have currently a room in University where we have a monitor for room conditions sensors, and we have actuators for the venetian blinds and lights.

If it is possible please answer the following questions suggesting the simplest possible settings only to show the concept of RL applied to energy in buildings (this is the goal of the project, the simplest the better, only to show RL as a replace of control systems).

Questions:
i)Would this library be suitable for controlling with rewards the venetian blinds and lights to reduce energy consumption and maintain comfort settings?

ii)How long is the learning curve, ie following the steps to install and configure how long does it take a developer to build a working simplest example ?

Thanks in advance.