Let's Make Robots!

Theo Jansen style robot leg: prototype #2 and #3

I started tinkering with "Theo Jansen legs" again (follow the link first if you don't know what that means).

I am prototyping in corrugated cardboard and trying to come up with a design that could be produced in large numbers (six at least) in plywood.


The left leg (prototype #2) has a "boxy" design. This one holds its shape better, but is so much wider than the other design. The right leg (#3) is eight pieces of cardboard laid flat on the table. It is very wobbly. I improved the lateral stiffness by adding strips of material perpendicular. Yes I used hot glue. The second video shows exactly what I am afraid of in the flat design: buckling in the joints.

The green marking indicates where the weight of the vehicle will rest on the leg. It is one of the four pivots in the parallelogram. The pink markings indicate where the two cranks connect from the crank shaft.

The lower two pictures illustrate another fundamental difference in the two designs. Design #3 cannot fold all the way flat because the linkages in the parallelogram touch each other. Cleverly S-shaped "bones" might circumvent this issue. More scissors and glue tomorrow and then on to the plywood!


update 30 dec:

video by Maneuver

update 1 jan 2009:

Found an interesting interactive animation of a "Theo Jansen Linkage". It lets you tweak the numbers and see the resulting movements.

Continued in my next blog. All pages tagged with "theo jansen" here.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

 interesting. I'm digging out the lego now to do a quick prototype

 / vzz-clck-"Maneuver"

Couldn't help myself, and made a mock-up with Lego just to see it in action.


(dont know how to embed video in comments )

Posted a pic with annotations in the original post as well...

/ vzz-clck-"Maneuver"

at least yours are walking! kinda

(and comments are pretty much the only place on LMR that do not accept embedded video)

am trying to use a 4-bar linkage to drive a vertical walll climbing robot for my mechatronics assignment but dnt know how to desing the linkage lengths for optimization of the motion and to prevent rocking in one position. i tried to use 8-legs with 4 in the frame n the other on the outside of the frame. HELP PLIS

Hello Sesame. I am not going to make your home work for you. But if you have something specific to ask about any of my projects, I would be very happy to answer you. Make sure you post your question on the page that describes the project.

Sesame; Read the Rules

I'm thinking about making a rideable walker myself.  My plan is to use 175mm bicycle cranksets as the crank arm on the Theo Jansen linkages... to give an idea of the scale.  I want to build in steel.

I have a prototype in popsicle sticks, and am starting to work on a 3D-printed plastic prototype.  My project is documented at http://boim.com/Jansen  This site has an animation of the walking cycle, source code to solve for linkage locations through the cycle, and .scad models for proposed 1:10 scale plastic linkages.

I have a concern that under load the leading "knee" joint can collapse, moving in towards the fixed pivot node.  This is the node that looks like a slightly hyper-extended knee when the foot is the the middle of it's ground contact phase.  My popsicle prototype will collapse, but it has very little out-of-plane support, and the "knee" can move out of plane to collapse towards the fixed node.

I have seen other walkers tend to use double linkages (wide).  I still have hope that in steel, I may not need to do this.  I am worried about of out-of-plane movement, and would love to hear suggestions about how to deal with this (in steel).

I guess people do LMR pages for incomplete projects, so I may start one soon...