BoT Project Biweekly Report
Subject: Beginning of the Project
Project Title: BotWheels
Report Number: 1
Date: October 12, 2024
Week Range: Weeks 1-2
1. Progress Overview
During this period, tasks were completed to lay solid foundations for the entire project. The focus was on distributing tasks related to research and structuring work with the repository. Linters for Python (Ruff) and a markdown linter were introduced, unifying the workflow of the project.
Setting up the local development environment for all team members was a main task, ensuring everyone could work on the project locally. Some team members committed to their tasks effectively, resulting in progress ahead of schedule in certain areas. However, delays were experienced in others.
The team successfully diagnosed and preliminarily implemented visualizations of the bot's environment states, exceeding weekly project assumptions. This is expected to positively impact the implementation of an effective bot and the development of the PPO and A2C/A3C algorithms. An increase in the number of researched topics can be observed in the docs repository, which are being integrated into the core repository on separate branches. Weekly status meetings are conducted every Wednesday at 20:00.
2. Team Member Contributions, Marks, and Linked Issues
| Team Member | Mark | Summary of Contributions | Linked Issues |
|---|---|---|---|
| Jakub Cisoń | 4.5 | Led the team and conducted research on the bot's gameplay in an online game, focusing on anti-cheat detection. Task distribution errors occurred, leading to some members not receiving assignments. The absence of a management system was identified and corrected, improving future task planning. Efforts will be made to better match task sizes for equitable workloads. | - Research security systems of Rocket League and Anti-Cheat statements |
| Kacper Drozdowski | 3.5 | Focused on understanding basic agent movement and producing a supporting document. A document about the A2C algorithm was expected for conversion to A3C, due October 14, 2024, but no committed work has been made as of October 13, 2024. | - Implement Movement Testing Script for Bot Actions - Research A2C to A3C transformation |
| Mateusz Gościniecki | 4 | Implemented basic bot behaviors and delivered documentation as a foundation for advanced implementations. Assigned a task to implement the basic PPO algorithm but reported illness may delay this delivery. | - Implement a Basic Bot That Drives Straight - Set Up Basic PPO Training |
| Anna Ivanytska | 2 | Tasked with implementing the basic DQN algorithm, but this was not delivered due to late communication of implementation issues. Prepared a detailed research document as groundwork for future implementation, but very simple. | - Initial Implementation of the DQN Algorithm - Documentation for the DQN Algorithm |
| Igor Malkovskiy | 4.5 | Implemented basic bot behaviors and delivered documentation for future use. Assigned a basic PPO algorithm task, but illness may delay delivery. | - Implement a Basic Bot That Drives Straight - Set Up Basic PPO Training |
| Dawid Mielewczyk | 5 | Demonstrated excellent preparation for project work, establishing a work culture, and setting up Discord bots for merge request notifications. Configured linters and established pipelines. | - Verify and Optimize Project Configuration for Best Practices - Familiarize with the PPO Algorithm - Issue #35 |
| Camille Nadir | 3 | Summarized knowledge about the PPO algorithm, it was delivered, but very late, so there was no time, to merge it till 14.10. Tasked with providing a detailed list of game environment states, which was also not completed. | - Familiarize with the PPO Algorithm - Create Documentation Entry Summarizing the PPO Algorithm |
| Michał Pryba | 5 | Showed strong commitment in researching the A3C algorithm and addressing implementation issues. Responsible for A3C algorithm implementation in the project. | - Initial Implementation of the A3C Algorithm - Investigate A3C Algorithm - only documentation for now |
| Wojciech Szamocki | 5 | Demonstrated reliability by diagnosing possibilities for visualizing agent states and the environment. Initiated work on the core repository based on personal initiative. | - Create visual mockup - Develop (Transparent if possible) Overlay/Window |
| Konrad Siemiątkowski | 2 | No task was assigned in the first week, which was the leader's responsibility. In the second week, a small task was assigned to improve the welcome page on GitHub docs, but this was not completed. | - Write a Welcome Page for Documentation |
| Michał Zarzycki | 3.5 | Tasked with delivering and examining the implementation of the PPO algorithm, but this was not delivered. Redid the previously assigned task of creating a test environment. | - Familiarize with the PPO Algorithm - Test PPO Integration with the Game |
3. Challenges and Problems Encountered
The person responsible for the initial implementation of the DQN algorithm did not fulfill their task, leading to the decision to abandon this approach for now. As an equivalent, documentation was prepared on this approach, which will serve as a foundation for future implementation if sufficient time and resources become available.
Some tasks are blocking the progress of other tasks, slowing down overall workflow and distribution. The engagement of group members is very varied, leading to communication problems and, as mentioned earlier, mutual blocking of tasks.
4. Plans for the Next Period (Weeks 3-4)
A better implementation strategy will be developed, and the distribution of human resources will be changed. Direct addressing of issues will be implemented to avoid blocking. A new task management system will be created based on the experience from the previous two weeks. A task planning system for the next few weeks will also be established. The A2C and PPO algorithms will be implemented, focusing on improving core functionalities.
5. Summary
In summary, the project started well but faced significant challenges during the first two weeks due to miscommunication regarding task assignments and varied engagement from team members. Plans for the next period aim to establish clearer roles and responsibilities to ensure smoother progress.