AnyOne out there using “RAMSIS-data” as general input (ie both motion and geometrical size) to your Models? If so, any hints?

Hi jokke

What kind of data export options exist? can the motion be exported?
Joints centers ?

Have a nice weekend

Best regards

Bonjour Sören!

I just found out that I can export the joint centers and joint angles from the RAMSIS data. I believe joint centers would be the best (easiest) choice. And, since you asked about them, my guess is that you think so too.

So, you have a suggestion?

Perhaps something along the lines…
Choosing some “major” joint centers, then do body segment scaling (“anthropometrics”), then drive the body (manikin) sort of like the “Drive successfully a model with the manual marker method” found on the wiki pages?

(Yeah, it was a nice weekend at Lac du Bourget, France. I hope you have some summer left in Ålborg!)

Bonjour jokke

Sounds good that the joint centers are available.
I think your approach sounds feasible, of course the joint angles are also tempting since it is a robust way of driving the motion. The drawback with the angles are that if the models have different joint topology this way of driving may introduce large errors in the extremities. Using the positions of the joints center this is avoided.

What kind of motion are you trying to do?

Best regards

No more summer seem to be left for Aalborg this year !


It will probably be a clutch pedal operation. Maybe a gear shift or a hand-brake operation. I do not know yet for which movement I have (external) force data but it will definitely be a vehicle driver. Thus, a slow and almost semi-static movement compared to what I am used to.

If I’ll do the clutch (and perhaps only need the lower body), do you recommend that I try the “GaitApp” first?

/jokke (guest researcher at INRETS, Lyon, France for the moment)

Hi Jokke. Version 4.1 is due soon and it will contain the facilities that were previously contained in GaitApplication2. In general, this version allows you to process over-determinate kinematics. i.e. too many markers, and also optimize the model parameters directly inside AnyBody.

Bonjour John!

v4.1 sounds really promising.