For anyone else who was curious and did not want to download an Excel file to read the table, parts will set you back $1300. Plus, who knows how much printing time.
I replayed in another comment, seems like it is way better to buy minipupper that is also servos based, it does not make sense spending so much on servos based robot, arguably, it is better to spend 1300+ building bot with actual motors Lorcan[1] alike (there others with 6 motors, if recall correctly).
There is also minipupper2[1] (I have minipupper1, waiting for version2), v2 has servos feedback (I think this first of kind in cheapest robots space), furthermore all minipupper1 hardware is also open-sourced on github[2], it is expected to be same for v2 version.
A Dingo is a shy (with humans), wild dog in Australia. https://en.wikipedia.org/wiki/Dingo. I saw once once, but it was about maybe 100-200 metres away.
If you're short of bandwidth or limited by a quota, uBlock Origin can be toggled easily between allowing or blocking all images. It's what I use to make my 750MiB per month GSM/4G connection last.
I am also curious about educated answers to this question, but I imagine taking a cue from millions of years of evolutionary pressure is a pretty solid bet.
8 legs are also well representated and going strong for millions of years, not to mention carcinisation being a meme at this point.
When it comes to robotics, my best guess would be cost and complexity. We also either need a better control interface or better self driving software to take advantage of the legs moving independently (I assume optimizing how the 8 legs should move to climb a rocky path for instance is just a computational nightmare to do realtime)
Readit News
https://github.com/222464/Lorcan-Mini
It's not quite as bad as it seems at first blush.
[1] https://www.kickstarter.com/projects/336477435/mini-pupper-2...
[2]https://github.com/mangdangroboticsclub
$529 - Pupper with Ri
$129 - Lidar
$148 - 3D Camera
(via https://news.ycombinator.com/item?id=36299315, but no comments there)
https://en.wikipedia.org/wiki/Dingo_ate_my_baby
---
Without reading the actual code, the structure of the system looks very clear, pleasant to understand. It's not a mess.
Once the expense of four has been invested is there a better number, say six or eight, to have?
I can imagine all sorts of advantages and disadvantages. They choose four, is that because four is best? Or because a dog has four?
When it comes to robotics, my best guess would be cost and complexity. We also either need a better control interface or better self driving software to take advantage of the legs moving independently (I assume optimizing how the 8 legs should move to climb a rocky path for instance is just a computational nightmare to do realtime)
Two is way hard. Six or more is also hard if you want to do it properly over rough terrain. Four is a sweet spot for human-scale robots.
But actual navigation is software related.