Experimenting with robots is too expensive. The software is free,
many designs use common Arduino and Raspberry Pi, but the
market is still dominated by proprietary "families of products".
Could some simple mechanical standards and 3D printing help out?
Could we make ROS compatible turtle-style robots for half or even less?
Or use the compatibility, while experimenting with other shapes?
Project static pages are: www.dodeck.net with details, history & inspriation, etc
github has a collection of STL files for disks and servo adapters. and just
enough software to move around: Phone to BLE to serial to servos.
A simple remote control for a Nordic BLE dev kit board, serial to Arduino.
A small Arduino project to move according to serial messages with angular and velocity.
The name comes from 'dodecagon' the almost round 12 sided polygon, plus a platform 'deck'.
I had a 6 sided robot long ago, but they used those horrible little drag it sideways (mechanum) wheels. All the JPL and NASA rovers use motors to steer the wheels directly, not dual-differential. And they can handle terrain on other planets, not just flat concrete floors. How could I do that much cheaper? Or at least not pick a geometry that prevented experiments like that. The turtle style dual-diff is the most common for education.
12 sides with 2,3,4,6 symmetry allows easy experiments with various shapes.
This looks like a good place to talk design.
The name comes from 'dodecagon' the almost round 12 sided polygon, plus a platform 'deck'.
I had a 6 sided robot long ago, but they used those horrible little drag it sideways (mechanum) wheels. All the JPL and NASA rovers use motors to steer the wheels directly, not dual-differential. And they can handle terrain on other planets, not just flat concrete floors. How could I do that much cheaper? Or at least not pick a geometry that prevented experiments like that. The turtle style dual-diff is the most common for education.
12 sides with 2,3,4,6 symmetry allows easy experiments with various shapes.