A Docker image that is uploaded to the Docker Hub and automatically downloaded from the cloud to a local computer, where it is deployed using Docker Compose (this solution is primarily suitable for development, taking into account the use of additional tools for preparation, configuration, and management)
A self-contained Python library that can automatically download additional dependencies and is currently being prepared for upload to the cloud for quick installation via Pip (currently only works locally)

Production #

(Planned for Week #6, when the full stack is stable and safe for hardware control)

Vibe Check #

Progress: Team is moving forward with hardware design integration and RL. The DQN agent is functional in simulation, and environment interfaces are maturing.

Roadblocks: Hardware integration and accurate reward shaping are pending. Limited real-time feedback testing due to incomplete API.

Team dynamics: Good communication. Everyone has clear tasks, but we need to ensure syncing frontend/backend progress with the ML environment timeline and documentation

General summary: All team members are very excited about the direction we are taking, especially the technologies we need to use and the future results

Weekly commitments #

Individual contribution of each participant #

Evgenii - Make PCB, box and tests for py
Artyom - Make CI/CD pipeline and docs
Anastasia - Read and write pipeline and data processing
Marat - Firmware optimization and tests
Petr - RL and Report

PCB

Plan for Next Week #

Add real-time hardware data processing
Finish deployment pipeline with automatic container builds and docs processing
Tune hyperparameters of RL agent
Finalize reward shaping logic
Expand unit and integration test coverage
Refactor entire code (Py, c++ and other)
Imporove assembly

Confirmation of the code’s operability #

We confirm that the code in the main branch:

In working condition.
Run via docker-compose (or another alternative described in the README.md).