Box weight included in GaitFullBody 1.5

[dstambolian]

@mkjung9980

Hi Damon,

In the 'ModelSetup.any' file there are all the marker definitions for the human segments. And they are using some special class templates for that purpose.

But for the markers regarding the box, you can't use that class template.

If you could see the tutorial (Making Things Move), then you know that you can use 'AnyKinDriverMarker' class to driver your box according to the related markers.

So it does not matter where you can define your code.

Best regards,
Moonki

Thanks Moonki

Possibely i am not looking in the correct location for the PosInterpol code.

Sincerly thanks
Damon

[dstambolian]

@mkjung9980

Hi Damon,

In the 'ModelSetup.any' file there are all the marker definitions for the human segments. And they are using some special class templates for that purpose.

But for the markers regarding the box, you can't use that class template.

If you could see the tutorial (Making Things Move), then you know that you can use 'AnyKinDriverMarker' class to driver your box according to the related markers.

So it does not matter where you can define your code.

Best regards,
Moonki

Hi Moonki
The box is now moving with the hands, thanks very much for your help to accomplish this.

[dstambolian]

Hi Moonki
The box is now moving with the hands :), thanks very much for your help to accomplish this. Here it is in youtube http://youtu.be/3LraFdNhG2Y

Now I’m trying to do the kinetic connection between the box to hands using AnyReacForce.

Q1 - Is there any lessons that are specific to AnyReacForce being used to connect hands to an object, or other helpful lessons?

Q2 - The initial location of the hand placement is always on the sides, but with each new trail the hands are not always in the same exact spot. So will AnyReacForce allow this flexibility in the location of hands on box?

Q3 - Would I only need AnyReacForce code as shown below, or do I need additional code for things like, nodes,

[SIZE=3]AnyReacForce <ObjectName> =
{
//Type = ;
//AnyKinMeasure &<Insert name0> = <Insert object reference (or full object definition)>; You can make any number of these objects!
};
[/SIZE]

Q4 - I tried running the Inverse Dynamic Model. I see the bones and muscles, but not the box. Do I need to add the box code into the muscle section of the main file?

Sincerely thanks
Damon

[mkjung9980]

Hi Damon,

Thanks for the nice video on YouTube.

Q1 - Is there any lessons that are specific to AnyReacForce being used to connect hands to an object, or other helpful lessons?
A1) There is no specific lessons for AnyReacForce.

Q2 - The initial location of the hand placement is always on the sides, but with each new trail the hands are not always in the same exact spot. So will AnyReacForce allow this flexibility in the location of hands on box?
A2) You don’t need to care the orientation of the hand.

Q3 - Would I only need AnyReacForce code as shown below, or do I need additional code for things like, nodes,
A3) You can just define AnyReacForce objects like this:

AnyReacForce LeftHandForceToBox = 
{
   AnyKinLinear lin = 
   {
       AnyRefFrame& ref1 = ... (somewhere).Hand;
       AnyRefFrame& ref2 = ... (somewhere).Box;
   };
   AnyKinRotational rot = 
   {
       Type = RotAxesAngles;
       AnyRefFrame& ref1 = ... (somewhere).Hand;
       AnyRefFrame& ref2 = ... (somewhere).Box;
   };
};

Q4 - I tried running the Inverse Dynamic Model. I see the bones and muscles, but not the box. Do I need to add the box code into the muscle section of the main file?
A4) You should include the reference of the box in your Model folder.

One more suggestion.
When I looked your video, the markers do not look fit very well. I recommend you to run the MotionAndParamterOptimization process again well.

Best regards,
Moonki

[dstambolian]

Hi Moonki

How do I reference the box in the Model Folder?

The markers are closer in this last run of MotionAndParamterOptimization, http://youtu.be/17DrYW4WmtY

Sincerely thanks
Damon

[mkjung9980]

Hi Damon,

You can just make the references of your box, marker drivers for the box in your model folder. Please read the following tutorial chapter about how to maker references:
http://www.anybodytech.com/fileadmin/AnyBody/Docs/Tutorials/chap7_Advanced_script_features/lesson1.html

If you want to know whether the weight of the box would be transferred to human, you can just try to run the inverse dynamics and try to see the shoulder joint reaction forces.

Best regards,
Moonki

[dstambolian]

Hi Moonki

I was able to load successfully the box in InverseDynamicModel by replicating the same code in MotionAndParameterOPtimization, and with inserting the new object name for MyBox. I figure I can use referencing later on to improve the code, once I know the model is working.

For now I’m getting this error described below when trying to run the model. Can you please help with what I should do about this error?

[SIZE=1]ERROR(OBJ.MCH.KIN2)[/SIZE] : [SIZE=1]C:/U…s/d…l/D…s/+…l/1…l/1…y/1_DamonBoxLiftngModel.main.any[/SIZE] : [SIZE=1]InverseDynamicStudy[/SIZE] : Model is kinematically over-constrained : Position analysis failed : 6 unsolvable constraint(s) found

Sincerely Thanks
Damon

[mkjung9980]

Hi Damon,

You can try to change the type of your constraints(markers) as soft constraints.
That might help.

Best regards,
Moonki

[dstambolian]

Hi Moonki

Here is a video showing the unresolved constraints error, http://youtu.be/SeJBPgcBCbI , I was able to correct the unresolved constraint(s) error by adding this code:

AnyKinEqType ANY_CTYPE = Soft;
InitialConditions.SolverType = KinSolOverDeterminate;
Kinematics.SolverType = KinSolOverDeterminate;

I found this method at: http://www.anybodytech.com/fileadmin/AnyBody/Docs/Tutorials/chapX_SolidWork2AnyBody/lesson2.html

[COLOR=magenta][COLOR=black]Then somehow this new error showed up about, Time, ‘t’. [/COLOR][/COLOR]

[SIZE=1]ERROR(OBJ1)[/SIZE] : [SIZE=1]C:/U…s/d…l/D…s/+…l/1…l/1…y/JointsAndDriversOptimized.any[/SIZE] : [SIZE=1]JntDriverTrunk[/SIZE] : Time, ‘t’, has an invalid value for this interpolation

[COLOR=black]Which I solved by changing the number 2 to a number [COLOR=black]6 in this below code. I think this just means it starting and stopping the motion 6 steps earlier. Now its loading and running in InverseDynamics, but it is taking a very long time to complete dynamic processing of the muscles. So, I’m going to try the same C3D file without the box and see if there is a difference.[/COLOR][/COLOR]

tStart=Main.TrialSpecificData.tStart+6Kinematics.ApproxVelAccPerturb;
tEnd=Main.TrialSpecificData.tEnd-6Kinematics.ApproxVelAccPerturb;

I found the relation to Time, ‘t’ at:
http://www.anybodytech.com/fileadmin/AnyBody/Docs/Tutorials/Making_things_move/lesson2.html

Q1 - For the mass moment of inertia, I’ve seen in the lessons the values for Jii = { ?, ?, ?}; are small, 1, 0.01, 0.0001, etc. I tried doing calculations using the link you provided, but my numbers are much larger, and I get a giant object on the screen. So I’m wondering if in the AB lessons if the three Jii numbers are being divided by the larger number. Example: jii = {500, 500, 250}; would be jii = {500, 500, 250}/5000; which would be jii = {0.1,0.1,0.05}; Could you please help with how calculations are done for AB a rectangular solid box?

Q2 - I had these below notices, is there something I can do to improve the code to address these notices?

Evaluating constants…
[SIZE=1]NOTICE(OBJ1)[/SIZE] : [SIZE=1]C:/U…s/d…l/D…s/+…l/1…l/1…y/EnvironmentAutoDetection.any[/SIZE] : [SIZE=1]ForcePlateType2_Origin_Z_value_message[/SIZE] : The older AMTI origin has its Z value as positive. So this value is converted automatically. Please refer to www.c3d.org/HTML/type21.htm.
[SIZE=1]NOTICE(OBJ1)[/SIZE] : [SIZE=1]C:/U…s/d…l/D…s/+…l/1…l/1…y/EnvironmentAutoDetection.any[/SIZE] : [SIZE=1]ForcePlateType2_Origin_Z_value_message[/SIZE] : The older AMTI origin has its Z value as positive. So this value is converted automatically. Please refer to www.c3d.org/HTML/type21.htm.
[SIZE=1]NOTICE(OBJ1)[/SIZE] : [SIZE=1]C:/U…s/d…l/D…s/+…l/1…l/1…y/EnvironmentAutoDetection.any[/SIZE] : [SIZE=1]ForcePlateType2_Origin_Z_value_message[/SIZE] : The older AMTI origin has its Z value as positive. So this value is converted automatically. Please refer to www.c3d.org/HTML/type21.htm.
[SIZE=1]NOTICE(OBJ1)[/SIZE] : [SIZE=1]C:/U…s/d…l/D…s/+…l/1…l/1…y/EnvironmentAutoDetection.any[/SIZE] : [SIZE=1]ForcePlateType2_Origin_Z_value_message[/SIZE] : The older AMTI origin has its Z value as positive. So this value is converted automatically. Please refer to www.c3d.org/HTML/type21.htm.
Configuring model…

Sincerely thanks
Damon

[dstambolian]

@dstambolian

Hi Moonki

I've tried several things and nothing has worked so far to solve the error. Below are two types of codes I've been testing.

Is either of these codes what should use to solve the error, and where should I put this code?

CType={Soft,Soft,Soft};

Or

[SIZE=3]InitialConditions.SolverType = KinSolOverDeterminate;

[/SIZE]
Sincerely thanks
Damon

HI Moonki, have you seen my latest post, and are just trying to deterime the solution?

Do you think it would help if you were to look at what I have done so far with the model, so you can have a better picture of what I have in the model and to provide recommendations to fix the errors?

Sincerly thanks
Damon

[mkjung9980]

HI Damon,

Now your questions became too complex for us to figure out at a glance.

I think you have to upload a zip file which contains your model.

Of course that model should work on our normal AMMR.

Please let us know the corresponding AMMR version related to your model.

Best regards,
Moonki

[dstambolian]

Hi Moonki

Here is all the files needed to work in AMMRV 1.5.1. There are 2 additional AB files that have some differences that need switched in 1.5.1. Also included is the CAD file for the box.

So attached is:
-Main
-EnviromentAutoDetection
-ModelSetup
-UnitBox

Thank you for looking at this.

Also still need to check if the box forces are translating to the arms.

Sincerely thanks
Damon

[dstambolian]

Hi Moonki

… Here is all the correct files needed to work in AMMRV 1.5.1.

There are 2 additional AB files that have some differences that need switched in 1.5.1. Also included is the CAD file for the box.

So attached is:
-Main
-EnviromentAutoDetection
-ModelSetup
-UnitBox

Thank you for looking at this.

Also still need to check if the box forces are translating to the arms.

Sincerely thanks
Damon

[mkjung9980]

Hi Damon,

We can’t manage individual modified versions of AMMR for individual questioners.

I would suggest you to use a kind of cloud service such as DropBox or GoogleDrive.

Then you can make a zip file which contains both your model and the modified AMMR.

If you can write the link for that shared model then we will try to download it.

Best regards,
Moonki

[dstambolian]

Hi Moonki

[SIZE=3][SIZE=2][COLOR=#000000]Were you able to access the model in google drive ok?[/SIZE][/COLOR][/SIZE]

[SIZE=2]https://drive.google.com/folderview?id=0ByKJo85n4mZOel9FMnd4d2YxTG8&usp=sharing
[/SIZE]
Sincerely
Damon

[dstambolian]

Hi Moonki

There are some other things id like to do to improve the model, that has to do with EMG. Should I start a new string for that or wait until you get done reviewing the model?

Sincerely thanks
Damon

[mkjung9980]

Hi Damon,

I would recommend you to start a new post with your new model.

And your google drive link requires me to log in using my account.

Maybe providing a kind of public link which Dropbox may support would be better.

Best regards,
Moonki

[dstambolian]

Hi Moonki

Thanks again for all your guidance to get the model to work, and teaching me about AB in the process.

Here is the link to dropbox. Hope this works better?https://www.dropbox.com/sh/swvv2ys9xdk8sdo/IetRVyBXsm

And, here is the link to the new thread for the University of Miami box lifting model, http://forum.anyscript.org/showthread.php?p=17826#post17826

Sincerely thanks
Damon

[dstambolian]

Hi Moonki

Here is the link to the new thread for the UM box lifting model, http://forum.anyscript.org/showthread.php?p=17826#post17826

Should I move the previous questions to the new thread, or will we complete these questions here in this thread?

Sincerely thanks
Damon

[dstambolian]

@mkjung9980

Hi Damon,

If you think that using a class template for box is too difficult for you,
then you can just define an AnySeg object with proper mass properties.

You can find the proper mass properties of a box here:
List of moments of inertia - Wikipedia

I attached a zip file which contains a unit sized box STL file.

So simply speaking, you can create your box very easily like this:

AnySeg MyBox = 
{
  Mass = ?; // Determine the mass as you want!
  Jii = {?, ?, ?}; // Determine the moments of inertia as you want!
 
  AnyDrawSTL DrawBox = 
  {
     FileName = "UnitBox.STL";
     RGB = {0, 1, 0};
     ScaleXYZ = {?, ?, ? }; // Determine the size as you want!
  };
 
  AnyRefNode MarkerNode1 = { ... };
  AnyRefNode MarkerNode2 = { ... };
};

I think that this way of creating a box will be much more easier to you rather than using a class template currently.

Of course you can upload your model if you want. But if your model is too complex then definitely it will take some time for us to review it.

I hope this may be helpful to you.

Best regards,
Moonki

Hi Moonki

Q-1) Is the moment of inertia for Jii in these units kg/m2 ?

Sincerely thanks
Damon