Week 1

Thinking of final project ideas is always the most daunting task. I was hard at work the other day, thinking of final project ideas, staring at a blank screen. The screen remained blank not only for lack of content, but also because of our cat. We got her in May and she’s a sweet cat, but very scared. A picture:

Figure 1

The issue with the cat is that she clearly wants our affection, but is too scared to receive it in almost every way. The only strategy so far has been play, using a long fishing rod with a feather at the end:

Figure 2

The issue with this strategy is it requires one of my hands for a long time, and it is impossible to come up with final project ideas with only one hand. So a better strategy may be something similar to The Blair Winch Project. The project uses 4 winches to constrain and move an orb in 3D space. I would replace the orb with a feather.

One issue with this approach is that at least one winch will always be lower than the orb/feather. This is a tripping hazard for the cat. Another issue is that the cat should be able to move the feather a bit. So, instead, I propose using 3 static winches to define a plane approximately parallel to the ground, and a fourth mobile winch for elevation:

Figure 3

The three horizontal cords should be chosen to be relatively inelastic, so that the moving winch remains at a relatively constant altitude and the control is easy to calculate. The vertical cord can be made elastic for added cat enjoyment. It also needs to be cat-safe.

In addition to the winches, cords and feather, this project will involve a joystick to control the position of the feather. The joystick will talk to the winch controller(s) over e.g. IP, so that we can use it when we are not at home. So that we can see the cat in that situation, the project will also involve a camera.

Because it’s 2025, the project should have a little bit of ML. The camera can be used to determine cat position, cat orientation, cat velocity, and instantaneous cat elongation coefficient (a second camera might help). Hopefully, a little neural network would be able to figure out how a loving owner moves the feather in response to the cat. The network will need training data, so part of the project will involve playing with the cat.

Vector Graphic

The first step is to create a vector representation of this project that doesn’t look like Figure 3. My attempt using Inkscape is shown below:

Figure 4