I’m having some trouble with a model I’m working on and I’m hoping you
could help. It’s a lower body model much like the Gait3D model, and as
a result I tried to mimick its constraints. However, for some reason
the kinematic analysis won’t run. The analysis ends at step 0.
My model is really based on a balance study where a person stands on a
tilting platform. I couldn’t get the constraints right with the full
model, which includes a platform attached to the foot. So I commented
out the platform definition portion of the model, which is in the
environment file, and made it as simple as possible so that (as a
starting point) it would be comparable with the Gait3D model. Even
with doing this however the model neglects to run. Can someone please
take a look at my model and help me figure out what I’m doing wrong.
Your help will be greatly appreciated.
When you change the whole set of drivers like you did it is very
likely that the kinematic won’t work directly. Especially when
driving with markers, that make the model more sensible to kinematic
errors than when it is driven directly by joints angles. That mean
you have to go again through the process of replacing the markers one
by one and keep the model running at each step by adjusting some
parameters on the markers or in the initial position.
But it seemed the previous version of your model was working fine,
didn’t it? Maybe you should consider to keep working with the
previous one, as I understood that the deadline for your project is
coming soon.
Don’t doubt asking if you have more questions.
Best regards,
Sylvain, AnyBody Support.
— In anyscript@yahoogroups.com, “Jerome” <jromeo_sc_forever@…>
wrote:
>
> Hi there,
>
> I’m having some trouble with a model I’m working on and I’m hoping
you
> could help. It’s a lower body model much like the Gait3D model, and
as
> a result I tried to mimick its constraints. However, for some reason
> the kinematic analysis won’t run. The analysis ends at step 0.
>
> My model is really based on a balance study where a person stands
on a
> tilting platform. I couldn’t get the constraints right with the full
> model, which includes a platform attached to the foot. So I
commented
> out the platform definition portion of the model, which is in the
> environment file, and made it as simple as possible so that (as a
> starting point) it would be comparable with the Gait3D model. Even
> with doing this however the model neglects to run. Can someone
please
> take a look at my model and help me figure out what I’m doing
wrong.
> Your help will be greatly appreciated.
>
> I’ve uploaded the model…named… ‘Ank001’.
>
> Thank you.
>
> Jerome
>
The original deadline has passed and I’m know trying to work towards
a new one, which again is only a few weeks out. You’re right, the
previous version of my model would run kinematically, however it
wasn’t a good representation of the actual motion collected in the
lab. That model was driven mostly in the medial-lateral direction (in
which there is virually no experimental motion), while the majority
of the actual motion is in the anterior-posterior diections. This
resulted in the output of the model being very noisy and
unrepresentative of the data collected.
On taking a look at the Gait3D model I realized that the model was
being driven in such a way that it would capture the true motion from
the lab activity. Realizing the significant motions in my lab setup
occured in the same directions relative to the model (since the
global co-ordinate system is a little different), I tried to mimic
the constraints in the gait model. I think I have done so
successfully, but con’t decifer why the model will not run. I think
I’m stuck until I can figure this out.
Jerome
— In anyscript@yahoogroups.com, “AnyBody Support” <support@…>
wrote:
>
> Hi Jerome,
>
> When you change the whole set of drivers like you did it is very
> likely that the kinematic won’t work directly. Especially when
> driving with markers, that make the model more sensible to
kinematic
> errors than when it is driven directly by joints angles. That mean
> you have to go again through the process of replacing the markers
one
> by one and keep the model running at each step by adjusting some
> parameters on the markers or in the initial position.
>
> But it seemed the previous version of your model was working fine,
> didn’t it? Maybe you should consider to keep working with the
> previous one, as I understood that the deadline for your project is
> coming soon.
>
> Don’t doubt asking if you have more questions.
>
> Best regards,
> Sylvain, AnyBody Support.
>
>
>
>
>
> — In anyscript@yahoogroups.com, “Jerome” <jromeo_sc_forever@>
> wrote:
> >
> > Hi there,
> >
> > I’m having some trouble with a model I’m working on and I’m
hoping
> you
> > could help. It’s a lower body model much like the Gait3D model,
and
> as
> > a result I tried to mimick its constraints. However, for some
reason
> > the kinematic analysis won’t run. The analysis ends at step 0.
> >
> > My model is really based on a balance study where a person stands
> on a
> > tilting platform. I couldn’t get the constraints right with the
full
> > model, which includes a platform attached to the foot. So I
> commented
> > out the platform definition portion of the model, which is in the
> > environment file, and made it as simple as possible so that (as a
> > starting point) it would be comparable with the Gait3D model. Even
> > with doing this however the model neglects to run. Can someone
> please
> > take a look at my model and help me figure out what I’m doing
> wrong.
> > Your help will be greatly appreciated.
> >
> > I’ve uploaded the model…named… ‘Ank001’.
> >
> > Thank you.
> >
> > Jerome
> >
>
It is right that the motion is mainly on the anterior-posterior
direction. But the previous version of the model was actually driven
on the anterior-posterior direction and not only on the medial-
lateral as you seem to think. May be you get confused because the
drivers are on the local coordinate system for each segments, and
they are different than the global one so it gets confusing sometimes.
Let me explain you in witch direction is driven the previous model:
the thorax id driven on the anterior-posterior translation (forward-
backward)
the pelvis is driven on the 3 translations (sideway/forward-
backward/upward-downward)
the hip flexion and knee flexion are also driven in the anterior-
posterior direction
the hip abduction is driven in the medial-lateral direction
(sideway)
the toe is driven upward-downward, this drive the ankle plantar
flexion
Matching the new set of drivers to your model as you want to do can
take some long time, even more if you are not very experienced in
that task. I did not look closely at the results when I tried to run
your model and I can’t tell about the noise you are talking about.
But considering that the model was actually driven in the good
direction, you may get similar results by using the drivers from the
gait3D. So my advice is really to go on with the previous version.
You can improve it by driving the rotation of the pelvis and the arms
that are still fixed at the moment.
Otherwise if you want to use the drivers from the gait3D, then it is
a matter of improving the initial position so that the system can
solve the kinematic equations (you can refer for that to the steps
described by John previously).
Best regards,
Sylvain, AnyBody Support.
— In anyscript@yahoogroups.com, “Jerome” <jromeo_sc_forever@…>
wrote:
>
> Hi Sylvian,
>
> The original deadline has passed and I’m know trying to work
towards
> a new one, which again is only a few weeks out. You’re right, the
> previous version of my model would run kinematically, however it
> wasn’t a good representation of the actual motion collected in the
> lab. That model was driven mostly in the medial-lateral direction
(in
> which there is virually no experimental motion), while the majority
> of the actual motion is in the anterior-posterior diections. This
> resulted in the output of the model being very noisy and
> unrepresentative of the data collected.
>
> On taking a look at the Gait3D model I realized that the model was
> being driven in such a way that it would capture the true motion
from
> the lab activity. Realizing the significant motions in my lab setup
> occured in the same directions relative to the model (since the
> global co-ordinate system is a little different), I tried to mimic
> the constraints in the gait model. I think I have done so
> successfully, but con’t decifer why the model will not run. I think
> I’m stuck until I can figure this out.
>
>
> Jerome
>
> — In anyscript@yahoogroups.com, “AnyBody Support” <support@>
> wrote:
> >
> > Hi Jerome,
> >
> > When you change the whole set of drivers like you did it is very
> > likely that the kinematic won’t work directly. Especially when
> > driving with markers, that make the model more sensible to
> kinematic
> > errors than when it is driven directly by joints angles. That
mean
> > you have to go again through the process of replacing the markers
> one
> > by one and keep the model running at each step by adjusting some
> > parameters on the markers or in the initial position.
> >
> > But it seemed the previous version of your model was working
fine,
> > didn’t it? Maybe you should consider to keep working with the
> > previous one, as I understood that the deadline for your project
is
> > coming soon.
> >
> > Don’t doubt asking if you have more questions.
> >
> > Best regards,
> > Sylvain, AnyBody Support.
> >
> >
> >
> >
> >
> > — In anyscript@yahoogroups.com, “Jerome” <jromeo_sc_forever@>
> > wrote:
> > >
> > > Hi there,
> > >
> > > I’m having some trouble with a model I’m working on and I’m
> hoping
> > you
> > > could help. It’s a lower body model much like the Gait3D model,
> and
> > as
> > > a result I tried to mimick its constraints. However, for some
> reason
> > > the kinematic analysis won’t run. The analysis ends at step 0.
> > >
> > > My model is really based on a balance study where a person
stands
> > on a
> > > tilting platform. I couldn’t get the constraints right with the
> full
> > > model, which includes a platform attached to the foot. So I
> > commented
> > > out the platform definition portion of the model, which is in
the
> > > environment file, and made it as simple as possible so that (as
a
> > > starting point) it would be comparable with the Gait3D model.
Even
> > > with doing this however the model neglects to run. Can someone
> > please
> > > take a look at my model and help me figure out what I’m doing
> > wrong.
> > > Your help will be greatly appreciated.
> > >
> > > I’ve uploaded the model…named… ‘Ank001’.
> > >
> > > Thank you.
> > >
> > > Jerome
> > >
> >
>