19 Feb

We always like to see when we can remove more lines of code than we add. The refactor mentioned in the blog post two weeks ago is finally done and merged, and in general refactoring is wrapping up. Hopefully new feature development (port arrays, then multipack devices) will start next week. A...

18 Feb

The goal detection vertical flying test: The goal detection horizontal flying test: The goal detection hovering flying test: Next step: Need to make a separate PID module for different situations/areas. Function will take in the PID for x,y, and input Improving the horizontal...

14 Feb

The main focus of this week was to familiarize with the use of brushless motors so that we can transition our system to use BLDC instead of regular DC motors. First we had to make sure that we can run it using an ESP32. In order to do that, we utilized Adafruit Servo Shield, which allows us to conne...

13 Feb

Testing the detection of the short range orange circle using Laptop's power: Testing the detection of the short range green square using Laptop's power: Testing the detection of the short range green triangle using Laptop's power: Testing the detection of the long range and big size gree...

11 Feb

Since most of this week has been a refactoring slog on the HDL compiler, I'll instead talk about the engineering for the source-measure unit (SMU, HDL here) and multimeter (HDL here) example design boards that were built last year, but only started to get brought up. As a recap, a SMU is a combi...

09 Feb

In this video, we use color threshold algorithm to detect the goal and output the center of the goal from terminal (Though it was not showing in the following video).

09 Feb

One video snapshot from our flying test after we debug it: After first, it performed pretty bad since the color we previous took was from conference room. But after we tweak parameters, it was better in the lab environment.

06 Feb

For Week 5 of Winter 2022, our team worked on the following: A new ML green ball detection algorithm (See Zhaoliang’s blog) A modularized blimp code where everything is not fused into one .py file. This will be easier to use whenever there are changes to be made. An easy way of finding the c...

04 Feb

These slides are presented in group meeting. Talking about the evolution from individual Reinforcement Learning towards multi-agent mean-field algorithms. Especially, one of emphasis is that we discussed two different strategies based on mean-field. In order to understand easily, we named them age...

04 Feb

I tested the color threshold algorithm on the live video from ESP32-cam: This is the demo with low resolution: This is the demo with high resolution: