BleWinrtDll VS ml-agents

- Comes with ready-to-use models and textures

To generate roads, I downloaded .gpx files from my Strava activities, then implemented a parser to take coordinates and elevations and build roads in-game using EasyRoads3D. These tools combined worked quite well but required some manual work. I had to go over the entire length of the road to smoothen it up (due to GPS noise in .gpx files). The terrain also didn’t always perfectly align with the road, so I had to manually modify it.

I figured that steering would be a good way to keep the user engaged in the gameplay, instead of just sitting on the bike and pedalling. With required steering, you need to be engaged to not drive off the road.

There are some solutions to virtual steering in other smart trainer games, but they are hardware-based - e.g., for Zwift, you need to buy their hardware controller. I really wanted to keep it as simple and inexpensive as possible, so I used a smartphone mounted to the handlebars as a steering controller.

I implemented native apps for both iOS and Android. They use the gyroscope to get the steering value and broadcast it as a BLE peripheral. Implementing Bluetooth in Unity was a little tricky as there is no built-in solution. The implementation depends on the platform the game is running on. It boils down to writing a native plugin for each platform you support.

For macOS, there is an open-source library written in Swift - https://github.com/codeinversion/sensors-swift. It already has the necessary services and characteristics implemented. I had to modify it a little and compile it into a Unity plugin.

For Windows, there also is a library - https://github.com/adabru/BleWinrtDll, but it wasn’t working too well. Because I targeted UWP, I was able to implement my own Bluetooth logic in C# and compile it directly in Unity.

Overall, it was a really fun project to work on. I’m particularly happy that I was able to release it starting from scratch in just a couple of months, and I hope to continue working on it.

Using the absolutely wonderful BLE dll for Unity (https://github.com/adabru/BleWinrtDll/)

Looking at the image on the BleWinrtDll GitHub page, my understanding is that when I subscribe to a device, I should be able to see the data under the “Subscribe” button, which is in a text field labeled “characteristic value” within the editor.

I am a beginner to using ML Agents to simulate an environment for DL i am trying to trial runs by tinkering through different values between the action space and keep encountering this issue when attempting to run a new trial. I've tried mlagents-learn --force and mlagents-learn --run-id=newtest but both prompt the same error message. Using linux, I am aware of a similar bug occuring in older versions (https://github.com/Unity-Technologies/ml-agents/issues/1505) but solutions didn't fix it.

The Github page for ML-Agents has a fairly straight forward example.

Very few commits in the last year and not many new features (https://github.com/Unity-Technologies/ml-agents/commits/develop)

and PR content: https://github.com/Unity-Technologies/ml-agents/commit/ed212103e451449bf84711a4a8f7bf11dfb1211a

Unity can build a stand-alone application or be used as a library. Javascript is deprecated, and Boo along with it although it was never really supported to begin with. Various types of machine learning are supported through the ML-Agent Toolkit and pretty well documented. The toolkit has a Python API but you should be careful about doing anything too unusual in Unity because the documentation tends to have a lot of dead-ends.

Op, I'd just pull the repo again to a new folder from https://github.com/Unity-Technologies/ml-agents (use SourceTree for simplicity if you don't know git).

My github issue on their documentation: https://github.com/Unity-Technologies/ml-agents/issues/5899