# Following Vicon data

Hej!

We need help with following data measured by a Vicon system. The
markers were attached on a paddle and we want to drive our model with
this measured motion. We have managed to define the points on the
paddle but how do we get the points to follow the data?
Should we just choose three points to follow in order to not get an
overdetermined problem?
We saw some example on your homepage but the you just drive a joint
(like the elbow) with measured data. How do we do with a paddle which
is free to move in space?
We have put a spherical joint between each hand and the paddle.

HÃ¤lsningar
Ingrid and Yann-Erick

Hi Ingrid and Yann-Erick

So if i understand your model correctly you have two arms connected to a
segment using two spherical joints and this segment should be driven by
marker data. I am in doubt if you also have markers for the arms, I the
following I will assume that this is not the case.

Each arm in the Body Models has the following dof

3 rotations in SC

3 rotations in GH

1 rotation in elbow

1 rotation in forearm

2 rotations in the wrist

So each arm has in total 10 dof that needs to be driven.

You are adding one segment and in this way adding six rigid body dof but the
two spherical joints each takes away three dof.

So the dof count looks like this

2*10 free rotations in both arms

6 dof from the new segment

-2*3 from the two spherical joints.

This means you will need to in total to drive 20 dof on the arms and the new
segment.

Here is a suggestion for these drivers you will need for your model:

2*3 for the SC joints

2*1 rotation in forearm (pronation/supination)

2*2 rotations in the wrist (flexion/extenstion abduction/adduction)

2*1 rotation of GH (abduction)

6 dof on the new segment so that it will be entirely driven by markers.

The summation gives 20 as it should.

A good place to “harvest” many of these drivers are in the Freeposture
model, please look at the JointAndDriver.any file.

I assume that you have at least three markers on the paddle otherwise it
will not be possible to drive it. To drive the paddle you will need to drive
it using all three coordinates from one of the markers and two from another
and one from the last one. These measures should be setup using the local

It may look something like this

AnyKinEqSimpleDriver Marker1={

AnykinLinear lin ={

Ref=0; //use the local coordinate system of the paddle

AnySeg &ref2=â€¦reference to marker one

};

DriverPos={0,0,0};

DriverVel={0,0,0};

Reaction.Type={Off,off,Off}

};

AnyKinEqSimpleDriver Marker2={

AnykinLinear lin ={

Ref=0; //use the local coordinate system of the paddle

AnySeg &ref2=â€¦reference to marker two

};

DriverPos={0,0};

DriverVel={0,0};

Reaction.Type={Off,Off}

MeasureOrganizer={1,2};

};

AnyKinEqSimpleDriver Marker3={

AnykinLinear lin ={

Ref=0; //use the local coordinate system of the paddle

AnySeg &ref2=â€¦reference to marker three

};

DriverPos={0};

DriverVel={0};

Reaction.Type={Off}

MeasureOrganizer={2};

};

This is of course just one option, there are several others, when selecting
between the individual dof it is important not trying to attempt driving the
length of the segment.

I would recommend to look at the JointAndDriver.any files of the GaitVaughan
and the WheelChairRancho models they are both using same concept of driving
selected dof from the markers, in order to avoid creating an over determined
system

It is a good idea always to keep a running model, here is a suggestion for a
procedure

1 so start by adding all the drivers from the free posture model

2 add the paddle and drive it using the markers, create a good guess
on the initial position and orientaion

3 Manipulate the mannequin file so that the hands end up close to
the paddle so that the constraints in the spherical joints are almost
fulfilled.

4 Create the spherical joints.

5 Remove the drivers for the elbow flexion, gh external rotation, gh
flexion.

6 Then hopefully the model will run kinematically

I hope this helps you move

Best regards

SÃ¸ren, AnyBody Support

From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of ydelval
Sent: 28 October 2008 14:21
To: anyscript@yahoogroups.com
Subject: [AnyScript] Following Vicon data

Hej!

We need help with following data measured by a Vicon system. The
markers were attached on a paddle and we want to drive our model with
this measured motion. We have managed to define the points on the
paddle but how do we get the points to follow the data?
Should we just choose three points to follow in order to not get an
overdetermined problem?
We saw some example on your homepage but the you just drive a joint
(like the elbow) with measured data. How do we do with a paddle which
is free to move in space?
We have put a spherical joint between each hand and the paddle.

HÃ¤lsningar
Ingrid and Yann-Erick

[Non-text portions of this message have been removed]

Hej,

We tried to follow your instructions and we used the GaitModel to
understand how the markers should fit in. Now we have a
MotionTRackerIclude and we can load the model but when we try to run
(Kinematic Analys) we get:
" ‘Study’ : Model is kinematically indeterminate : Position
analysis failed : 244 independent constraints and 259 unknowns"
We have used 3 coordinates for the first marker, 2 for the second and
1 for the last.
Do you have any good idea how to solve this problem?

Ingrid and Yann-Erick

Hi Ingrid og Yann-Erick

The error message suggests that you are missing drivers for in total 259-244
= 15 dof.

First of all make sure you have the model running first before adding the
paddle, this you do by driving it by drivers on the joints angles directly.
These drivers can be taken from the free posture model.

Once the model is running then add the paddle segment and the free floating
markers.

If now trying to run the model it will say that you need to driver six dof
more, this is due to the extra paddle seg.

The drivers for the paddle now needs to be added, like you suggest 3 from
one, 2 from the second and one from the third. When applying these drivers
make sure to use the local coordinate system of the paddle.

To make it easier you can also start by doing a model with only the paddle,
no human model; once this is running you can connect the human model to the

Hope this helps you move on, otherwise please write again.

Best regards

SÃ¸ren, AnyBody Support

From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of ydelval
Sent: 12 November 2008 14:22
To: anyscript@yahoogroups.com
Subject: [AnyScript] Re: Following Vicon data

Hej,

We tried to follow your instructions and we used the GaitModel to
understand how the markers should fit in. Now we have a
MotionTRackerIclude and we can load the model but when we try to run
(Kinematic Analys) we get:
" ‘Study’ : Model is kinematically indeterminate : Position
analysis failed : 244 independent constraints and 259 unknowns"
We have used 3 coordinates for the first marker, 2 for the second and
1 for the last.
Do you have any good idea how to solve this problem?

Ingrid and Yann-Erick

[Non-text portions of this message have been removed]