Last week in robotics I have mainly focussed on writing more of my research report on embedded systems, looking at Boston Dynamics’ robot dog names Spot, and more specifically looking at its ability to map its environment in real-time and autonomously navigate. I started the assignment this week by looking further into the components and developing my base understanding, while getting an outline on how I could write the final report. I learnt a lot early on about how autonomous navigation works, understanding Simultaneous Localisation and Mapping (SLAM), and the parts that go into it such as LiDAR. This is a system that I am quite interested in, and I got to learn a good deal about how the system works. A lot of my underlying knowledge on SLAM algorithms came from watching the YouTube video by MATLAB titled “Understanding SLAM using pose graph optimization: Autonomous navigation, Part 3”. This video gave me a lot of insight about how SLAM works, and why it is necessary. It talked about how the robot must estimate its pose from how far it’s travelled, and keep taking live feedback from LiDAR sensors about where obstacles are relative to the robot. It also discussed things such as how the estimated pose of an autonomous robot may have even a slight difference from its real pose, it can cause a map of its environment to be disjointed, and while each map of the obstacles may be correct relative to its pose, the overall map will be out of place.

MATLAB. (2020, July 8). Understanding slam using pose graph optimization: Autonomous navigation, Part 3 [Video]. YouTube. https://www.youtube.com/watch?v=saVZtgPyyJQ at 4:32.

This type of assignment was fun to do, as the topic of it was very much up to what I wanted to research, with some guidelines. This freedom allowed me to develop my knowledge on a topic within robotics which I am genuinely interested, and on a company making amazing new robots.

I also had the Lockheed Martin CyberQuest this weekend. I was in a team with other people in my class, one of which was doing robotics and the other did web dev. I went into this competition very under-confident, as the only practice content I had was some CyberSec courses on Grok, which were aimed for people in year 5. Completing those were easy, however the actual competition was much harder. Our team struggled through, and although it seemed near the end our team did horribly (with 65 out of 345 points), it turns out we were only 12 points off placing, and 20 points off tying for first place. We also got informed by out coach that we were holding third place for some time during the two hour challenge. This competition, although stressful, was quite fun and I look forward to an opportunity to participate again next year.