Development blog for TORCS project.
by Oliver
On Monday morning, the team met up to record the final video. Kerem brought his Fujifilm X-T2, so we were able to record the video in 4k 30fps, allowing us to capture professional quality footage. The team booked a group workroom to record the video and took turns filming each others segments. After plenty of outtakes, everyone recorded a flawless cut, and with that the recording was complete! The camera captured the video, and a mobile phone to capture the audio. Once both were recorded, we used DaVinci Resolve to edit and cut the video, clipping it with the footage of the fastest lap video. The video was exactly 3 minutes, meeting the requirements perfectly. Complete with background music, captions, and an introduction, the video was professional quality.
After ensuring our team had met all of the project requirements, both functional and non-functional, updating our timesheets, and a final version of the risk register, we were able to submit our final project. This marked the end of our 12 weeks working as a team, and over this time we have significantly improved in our collaboration skills, and our ability to learn as we develop. Having never implemented reinforcement learning models before, this project pushed our team to learn, research and implement models that we had no prior experience with, but our organisation and planning allowed us to do so without any significant issues.
The complexity of having to learn as we developed was helped by IBM providing us with all of the necessary tools to complete the project. They gave us access to IBM Granite, and later on IBM Bob, their LLMs which we were able to use to break down, understand, and develop code for our solution. These alone would have been helpful, but in combination with IBM SkillsBuild courses, such as AI for Software Development, or Prompt Engineering, meant that we were able to hone our skills using these models. This made the development and implementaiton of our AI racer much smoother than it would have been without these tools. This experience with IBM has been very insightful, into what will be like to work on similar projects professionally, but also been incredibly fun to work on throughout - our team has grown close and everyone has become good friends.
Our recommended course: https://skills.yourlearning.ibm.com/activity/ALM-COURSE_4058914
tags: