It’s been less than a week since we launched our Kickstarter and we are 72% funded. This is mindblowing! Thank you everyone for donating, for spreading the word, and for believing in us.
We’ve received a couple of requests about the code repository and realized we haven’t ever publicly shared the link. Presently the central repository is a collaborative sandbox and full of experimental code, we will be separating a release branch out soon. In the meantime, the link for the curious is here.
We’ve also had generous offers from remote developers who want to help. Thank you! We’ll be figuring out the right way to incorporate public contributions as we clean up the repository, stay tuned.
Heightmapped terrain is now supported in the simulator. This will let us simulate much more complex and interesting situations than we could previously. Here’s a short demo of what that looks like with Stompy walking blindly forward. The ground is red in honor of the Mars landing.
The blue lines are also a new feature that help us visually gauge the stability of a gait. The polygon on the ground is the support polygon. The blue line pointing down from the center of the chassis indicates the acceleration of the center of mass of the robot. Where the projection of the center of mass meets the ground is roughly the center of pressure. The closer the center of pressure is to the center of the support polygon, the more stable the gait.
As most of you will have seen on our kickstarter page, we’ve received all waterjet parts for the full scale prototype leg and test fitted them together (everything fits). Welding starts on Saturday.
More updates soon, in the meantime enjoy this striking demonstration of leg scale that one of the team members shared on instagram (sorry for the filters… still a great picture!). For reference, Matt is 6 feet (1.83m) tall.
Thanks for reading, and for spreading the word. You are making this all possible.
Hi all, things looking great. I think what you are doing is fantastic.
what is the simulation software used in the video above?
Have you seen this? It is created in sketchup and simulated with sketchyphysics. I didn’t make it but I set it up on my PC with two logitech joysticks to drive it – really fun. The code controlling it is all rubyscript and amazingly simple: http://www.youtube.com/watch?v=M44W7wVR_l0&feature=fvsr
This too: http://www.youtube.com/watch?v=CD2V8GFqk_Y
Haven’t seen that video before! Didn’t know about the sketchyphysics plugin, thanks for sharing!
The simulator is based around the Open Dynamics Engine (ODE) and written in Python. Pygame and pyOpenGL handles the rendering. This simulation environment was written for the class and has not been branched as a separate project, but is available in the publicly linked codebase. Most of the critical files are in the “SimulationKit” directory.
The neat thing about Sketchyphysics is the ability to include control systems (joysticks etc), coupled with logic (if this then this else this). However it is a bit glitchy.
The software you have written is really impressive – not only are you advancing robotics but physics simulation as well.
I love the project