# My introduction & question about ModelInfo Operation for DOF analysis

Hello everyone!

This is my first question to this tech group.
Let me introduce myself.

My name is Moonki Jung, a PhD course student in Seoul National
Univeristy, South Korea.
My major is mechanical engineering, and my main research area were

Nowdays I start to do research on the field of biomechanics.
I’m interested in the application of biomechanical analysis program to
Because most of the existing CAD systems mainly focus on the design of
the products without considerations of users.
Although some CAD systems have DHM(Digital Human Modeling) functions,
their evaluation functions are not as good as those of Anybody.
In future, I think that the product design system like CAD must have
the powerful functions of the evaluation of biomechanical aspects of
the product.

So, I would like to implement an integrated system of CAD and
biomechanical analysis softwares like Anybody.
Because I’m not goot at the biomechanics,
I hope to know how to use the biomechanical analysis softwares like
Anybody well to improve my research.
This is my current research plan.

I’m not familiar with Anybody program. I’m trying to use this program.
very elementary to you.

1. To determine whether the model is kinematically determinate, we
use the ModelInfo operation. According to the tutorial, the total number
of rigid body DOF should be same as the sum of constraints from
joints, drivers and ‘ohters’ constraints.

[Total number of rigid body DOF= Total number of constraints(Joints,
Drivers, Others)]

This criteria performs an analysis on the whole model.
If the number of ‘Drivers’ are insufficient, the model would be
kinematically indeterminate. When we use the Anybody program to
design a sport equipment, we should define joints where the human and
product have contacts. And we should define drivers there properly.
But the designer would make mistake to forget to set drivers there,
then the ModelInfo analysis tells us that this model is kinematically
indeterminate. Can the Anybody program tell users where we should set
proper drivers to perform an analysis well automaically?

1. When we try to analyze very complex model like human-product
integrated model, we should set additional joints, drivers and
constraints. But I think that these user-defined constraints may
have conflicts with existing constraints in the default human model,
like ‘others’ constraints. Then we should eliminate some constraints.
Then, by what strategy should we eliminate constraints or change DOFs
of joints? Are there any principles to cope with these situations?

My questions may be too abstract to this tech group. If then, I

Hi Moonki Jung,

Welcome to the group.

When the system warns you about missing constraint it cannot tells
you where there are missing. So it is always good to keep track of
the drivers you modify, add or remove because the missing drivers
will logically be related to those changes.
Also to avoid missing drivers or to locate them, it is useful to
count them. There must always be as much constraints as dof. So if
for example you add 3 drivers to lock the hand on some mechanical
device, then you have remove 3 corresponding drivers (usually on the
wrist or elbow or shoulder), no less and no more.
Drawing some quick sketches also helps a lot to have an over view of
the drivers and see where there are missing (or redundant, there can
also be too much driver).

The recommendations are basically the same. When adding devices in
the body you are right that you will add joints, add and remove
drivers. You have to determine by yourself what are the dof allowed
or not by the device. For this my advice is once again to make some
sketches. It is a simple but efficient strategy. Of course all this
require a preliminary knowledge of the human model structure: which
are the dof of each modelled joint (3 rotations on the shoulder, 2 on
the elbow, etcâ€¦).
And the principle of the balance of the constraints always applies:
if you add a driver somewhere then you have to remove another
somewhere to keep equal number of constraints and dof.

Best regards,
Sylvain, AnyBody Support.

— In anyscript@yahoogroups.com, “mkjung99” <mkjung99@…> wrote:
>
> Hello everyone!
>
> This is my first question to this tech group.
> Let me introduce myself.
>
> My name is Moonki Jung, a PhD course student in Seoul National
> Univeristy, South Korea.
> My major is mechanical engineering, and my main research area were
> CAD(Computer-Aided Design) and computational geometry.
>
> Nowdays I start to do research on the field of biomechanics.
> I’m interested in the application of biomechanical analysis program
to
> Because most of the existing CAD systems mainly focus on the design
of
> the products without considerations of users.
> Although some CAD systems have DHM(Digital Human Modeling)
functions,
> their evaluation functions are not as good as those of Anybody.
> In future, I think that the product design system like CAD must have
> the powerful functions of the evaluation of biomechanical aspects of
> the product.
>
> So, I would like to implement an integrated system of CAD and
> biomechanical analysis softwares like Anybody.
> Because I’m not goot at the biomechanics,
> I hope to know how to use the biomechanical analysis softwares like
> Anybody well to improve my research.
> This is my current research plan.
>
> I’m not familiar with Anybody program. I’m trying to use this
program.
> very elementary to you.
>
> 1) To determine whether the model is kinematically determinate, we
> use the ModelInfo operation. According to the tutorial, the total
number
> of rigid body DOF should be same as the sum of constraints from
> joints, drivers and ‘ohters’ constraints.
>
> [Total number of rigid body DOF= Total number of constraints(Joints,
> Drivers, Others)]
>
> This criteria performs an analysis on the whole model.
> If the number of ‘Drivers’ are insufficient, the model would be
> kinematically indeterminate. When we use the Anybody program to
> design a sport equipment, we should define joints where the human
and
> product have contacts. And we should define drivers there properly.
> But the designer would make mistake to forget to set drivers there,
> then the ModelInfo analysis tells us that this model is
kinematically
> indeterminate. Can the Anybody program tell users where we should
set
> proper drivers to perform an analysis well automaically?
>
> 2) When we try to analyze very complex model like human-product
> integrated model, we should set additional joints, drivers and
> constraints. But I think that these user-defined constraints may
> have conflicts with existing constraints in the default human model,
> like ‘others’ constraints. Then we should eliminate some
constraints.
> Then, by what strategy should we eliminate constraints or change
DOFs
> of joints? Are there any principles to cope with these situations?
>
> My questions may be too abstract to this tech group. If then, I
>

I’m sorry for my late replying.

1. You said that the anybody system cannot tell us where the missing
drivers should be added. Then, do you have the plan to implement this
function in future?
And, do you think that research on implementing this function has an
academic meaning from the viewpoint of improving user’s convenience?

situations of missing or redundant drivers(constraints). Are there
some examples for these situations in this site? I’d like to get some
examples to solve.

3. In my lab, there are 2 copies of anybody. We want to use it in
Windows Vista operating system. But it doesn’t work well. When can I
use Anybody in Windows Vista OS?

— In anyscript@yahoogroups.com, “AnyBody Support” <support@…> wrote:
>
> Hi Moonki Jung,
>
> Welcome to the group.
>
> 1)
> When the system warns you about missing constraint it cannot tells
> you where there are missing. So it is always good to keep track of
> the drivers you modify, add or remove because the missing drivers
> will logically be related to those changes.
> Also to avoid missing drivers or to locate them, it is useful to
> count them. There must always be as much constraints as dof. So if
> for example you add 3 drivers to lock the hand on some mechanical
> device, then you have remove 3 corresponding drivers (usually on the
> wrist or elbow or shoulder), no less and no more.
> Drawing some quick sketches also helps a lot to have an over view of
> the drivers and see where there are missing (or redundant, there can
> also be too much driver).
>
> 2)
> The recommendations are basically the same. When adding devices in
> the body you are right that you will add joints, add and remove
> drivers. You have to determine by yourself what are the dof allowed
> or not by the device. For this my advice is once again to make some
> sketches. It is a simple but efficient strategy. Of course all this
> require a preliminary knowledge of the human model structure: which
> are the dof of each modelled joint (3 rotations on the shoulder, 2 on
> the elbow, etc�).
> And the principle of the balance of the constraints always applies:
> if you add a driver somewhere then you have to remove another
> somewhere to keep equal number of constraints and dof.
>
> Best regards,
> Sylvain, AnyBody Support.
>
>
>
>
> — In anyscript@yahoogroups.com, “mkjung99” <mkjung99@> wrote:
> >
> > Hello everyone!
> >
> > This is my first question to this tech group.
> > Let me introduce myself.
> >
> > My name is Moonki Jung, a PhD course student in Seoul National
> > Univeristy, South Korea.
> > My major is mechanical engineering, and my main research area were
> > CAD(Computer-Aided Design) and computational geometry.
> >
> > Nowdays I start to do research on the field of biomechanics.
> > I’m interested in the application of biomechanical analysis program
> to
> > Because most of the existing CAD systems mainly focus on the design
> of
> > the products without considerations of users.
> > Although some CAD systems have DHM(Digital Human Modeling)
> functions,
> > their evaluation functions are not as good as those of Anybody.
> > In future, I think that the product design system like CAD must have
> > the powerful functions of the evaluation of biomechanical aspects of
> > the product.
> >
> > So, I would like to implement an integrated system of CAD and
> > biomechanical analysis softwares like Anybody.
> > Because I’m not goot at the biomechanics,
> > I hope to know how to use the biomechanical analysis softwares like
> > Anybody well to improve my research.
> > This is my current research plan.
> >
> > I’m not familiar with Anybody program. I’m trying to use this
> program.
> > very elementary to you.
> >
> > 1) To determine whether the model is kinematically determinate, we
> > use the ModelInfo operation. According to the tutorial, the total
> number
> > of rigid body DOF should be same as the sum of constraints from
> > joints, drivers and ‘ohters’ constraints.
> >
> > [Total number of rigid body DOF= Total number of constraints(Joints,
> > Drivers, Others)]
> >
> > This criteria performs an analysis on the whole model.
> > If the number of ‘Drivers’ are insufficient, the model would be
> > kinematically indeterminate. When we use the Anybody program to
> > design a sport equipment, we should define joints where the human
> and
> > product have contacts. And we should define drivers there properly.
> > But the designer would make mistake to forget to set drivers there,
> > then the ModelInfo analysis tells us that this model is
> kinematically
> > indeterminate. Can the Anybody program tell users where we should
> set
> > proper drivers to perform an analysis well automaically?
> >
> > 2) When we try to analyze very complex model like human-product
> > integrated model, we should set additional joints, drivers and
> > constraints. But I think that these user-defined constraints may
> > have conflicts with existing constraints in the default human model,
> > like ‘others’ constraints. Then we should eliminate some
> constraints.
> > Then, by what strategy should we eliminate constraints or change
> DOFs
> > of joints? Are there any principles to cope with these situations?
> >
> > My questions may be too abstract to this tech group. If then, I
> >
>

Hi Moonki Jung,

We do not plan to implement a function that can tell you exactly where the
drivers are missing. But we are planning to implement in the future to a new
kinematic method that will be able to handle kinematically overdetermined
systems. We hope that with this new method it will be easier to spot the
missing drivers, for example we are thinking of using a set of “back up”
drivers that would be used if some driver is missing. So if one of the
segments is in his back up position that would mean the driver is missing.

Of course it would be an improvement for user convenience to have that kind
of facility pointing directly to where the drivers are missing. Do you plan
to set up a research project on this in your university?

We only provide working models on our web site and on the repository. But
you can easily set up a model for practice. A simple way is to start with
the standing model or the free posture model, create some nodes on the
GlobalRef and try to hook the feet or hands or pelvis to it. Then next step
can be to hook the hands for example to a moving segment so the arm will
follow it. It is a good exercise to do.

AnyBody is already running on Vista. However some users are experimenting
troubles. Please write to support@anybodytech.com and give us more details
about the problems you are having with Vista so we can solve it faster.

Best regards,

Sylvain, AnyBody Support.

From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of mkjung99
Sent: 14. august 2008 08:50
To: anyscript@yahoogroups.com
Subject: [AnyScript] Re: Re: My introduction & question about ModelInfo
Operation for DOF analysis

I’m sorry for my late replying.

1. You said that the anybody system cannot tell us where the missing
drivers should be added. Then, do you have the plan to implement this
function in future?
And, do you think that research on implementing this function has an
academic meaning from the viewpoint of improving user’s convenience?

situations of missing or redundant drivers(constraints). Are there
some examples for these situations in this site? I’d like to get some
examples to solve.

3. In my lab, there are 2 copies of anybody. We want to use it in
Windows Vista operating system. But it doesn’t work well. When can I
use Anybody in Windows Vista OS?

— In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
“AnyBody Support” <support@…> wrote:
>
> Hi Moonki Jung,
>
> Welcome to the group.
>
> 1)
> When the system warns you about missing constraint it cannot tells
> you where there are missing. So it is always good to keep track of
> the drivers you modify, add or remove because the missing drivers
> will logically be related to those changes.
> Also to avoid missing drivers or to locate them, it is useful to
> count them. There must always be as much constraints as dof. So if
> for example you add 3 drivers to lock the hand on some mechanical
> device, then you have remove 3 corresponding drivers (usually on the
> wrist or elbow or shoulder), no less and no more.
> Drawing some quick sketches also helps a lot to have an over view of
> the drivers and see where there are missing (or redundant, there can
> also be too much driver).
>
> 2)
> The recommendations are basically the same. When adding devices in
> the body you are right that you will add joints, add and remove
> drivers. You have to determine by yourself what are the dof allowed
> or not by the device. For this my advice is once again to make some
> sketches. It is a simple but efficient strategy. Of course all this
> require a preliminary knowledge of the human model structure: which
> are the dof of each modelled joint (3 rotations on the shoulder, 2 on
> the elbow, etc�).
> And the principle of the balance of the constraints always applies:
> if you add a driver somewhere then you have to remove another
> somewhere to keep equal number of constraints and dof.
>
> Best regards,
> Sylvain, AnyBody Support.
>
>
>
>
> — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
“mkjung99” <mkjung99@> wrote:
> >
> > Hello everyone!
> >
> > This is my first question to this tech group.
> > Let me introduce myself.
> >
> > My name is Moonki Jung, a PhD course student in Seoul National
> > Univeristy, South Korea.
> > My major is mechanical engineering, and my main research area were
> > CAD(Computer-Aided Design) and computational geometry.
> >
> > Nowdays I start to do research on the field of biomechanics.
> > I’m interested in the application of biomechanical analysis program
> to
> > Because most of the existing CAD systems mainly focus on the design
> of
> > the products without considerations of users.
> > Although some CAD systems have DHM(Digital Human Modeling)
> functions,
> > their evaluation functions are not as good as those of Anybody.
> > In future, I think that the product design system like CAD must have
> > the powerful functions of the evaluation of biomechanical aspects of
> > the product.
> >
> > So, I would like to implement an integrated system of CAD and
> > biomechanical analysis softwares like Anybody.
> > Because I’m not goot at the biomechanics,
> > I hope to know how to use the biomechanical analysis softwares like
> > Anybody well to improve my research.
> > This is my current research plan.
> >
> > I’m not familiar with Anybody program. I’m trying to use this
> program.
> > very elementary to you.
> >
> > 1) To determine whether the model is kinematically determinate, we
> > use the ModelInfo operation. According to the tutorial, the total
> number
> > of rigid body DOF should be same as the sum of constraints from
> > joints, drivers and ‘ohters’ constraints.
> >
> > [Total number of rigid body DOF= Total number of constraints(Joints,
> > Drivers, Others)]
> >
> > This criteria performs an analysis on the whole model.
> > If the number of ‘Drivers’ are insufficient, the model would be
> > kinematically indeterminate. When we use the Anybody program to
> > design a sport equipment, we should define joints where the human
> and
> > product have contacts. And we should define drivers there properly.
> > But the designer would make mistake to forget to set drivers there,
> > then the ModelInfo analysis tells us that this model is
> kinematically
> > indeterminate. Can the Anybody program tell users where we should
> set
> > proper drivers to perform an analysis well automaically?
> >
> > 2) When we try to analyze very complex model like human-product
> > integrated model, we should set additional joints, drivers and
> > constraints. But I think that these user-defined constraints may
> > have conflicts with existing constraints in the default human model,
> > like ‘others’ constraints. Then we should eliminate some
> constraints.
> > Then, by what strategy should we eliminate constraints or change
> DOFs
> > of joints? Are there any principles to cope with these situations?
> >
> > My questions may be too abstract to this tech group. If then, I