OpenSim crashes during ID-Analyses

[Matthias Kunze]

Hello everybody,

my inverse dynamic-analyses still crashes sometimes...
Especially if I choose my time-Range from 0 to 9 or similar. But not always on erery process. Why does it happens? Is there still a bug for the Start-time at 0 ? (the negative start at -1.16529e-011)

I take the same parameters for the several runs...
So the setup-files couldn't be the reason I think. (names, files etc conform...)

Thanks for help.

Matthias

[Matthias Kunze]

Maybe I got the problem by a process of elimination:
If I choose exact the absolut ending time of my recorded data as the end of the time range in the Inv. Dyn. Tool... the system crash. So I think the reason wasn't the start time = 0. Now I can run the tool with 0 to 8s (the data reach form 0 to 9.999s).

Maybe the time range need an advice for the users
("Don't choose same End-time...") in the next version.

Matthias

[Sylvia Wan]

Hi,
I have a similar problem where OpenSim crashed during ID.
I have read through the forum and tried several things but it is still no better.
I have tried changing Debug value to 3; changing the time range so that it is later than the start time and earlier than the finishing time in my data file; creating xml files from GUI as well as modifying xml files using Notepad ++; and checking that there are no gaps in the grf file. I have also compared my setting file with the gait2354 ID setting file and there were no significant differences.

In summary, there are two scenarios when ID crashed:
1. When I loaded previously saved setting, a message called ‘Java™ Platform SE Binary’ came up and said Java has stopped working before crashing.
2. If I do not load setting but load the relevant files and define external force set one by one, OpenSim is able to run ID for few seconds before crashing without giving any message.

In both scenarios, I didn’t get Stack Trace.

I wonder if it is the setting file that is causing the problem or if it is the model. I am using a modified lower body model with the torso deleted from Arnoadetal2010 model. However, I only deleted the relative bodies and muscles from the original osim file. So I don’t think the problems lie on the model.
Lastly, I am wondering if the IK and ID tools can generate IK_task.xml and Setup_IK.xml; as well as setup_InverseDynamics.xml on its own in the GUI? What I did was copying files from gait2392 and gait2354 and used them as a template, then changing the relevant parts to input my data. Where necessary, I overwrite the file by saving settings in the IK and ID tools. I just wonder how everyone else prepared their setting files. I would like to know if this is the best way to do it.
Many thanks!
Sylvia

[Sylvia Wan]

Hi,

I think my problem was solved by writing a tag for loading grf.mot in long. ie. <datafile> C:\Drivers\ApplicationPanel\OpenSim 2.0\Project\(8)ID\subject01_Ds_Walk11_grf.mot </datafile>

However, I am still interested to know how everyone created their setting files for IK and ID. Do you create them purly using the GUI or do you modify from existing example such as gait2392 in programming environment?

Also, did your IK results saved automatically? Mine did not saved even the <output_motion_file> tag specified a name for the output file. I had to right click on the trial name in the navigator to save the results into a mot file manually.

Would love to hear from anyone.

Sylvia

[Matthias Kunze]

Hi Sylvia,

I tested a lot of settings and always use the GUI to create the .xml-files for the ID and IK. (For Scale I look at the example-files and rebuilt my own files with Notepad.
Be sure that you have all files in the same directory and have all consistent names during different analyzis.

My IK results didn't save automatically, too. If I close the model a message helped to save the motion or - like your way - I saved the motion by a right klick.

Hope this helps.

Matthias

[Mohan Krishna Kallakanti]

Hi all,

I have similar problem. I've tried all the stuff discussed above. Can any body help me if I provided required files?

Thanks,
Mohan

[Sylvia Wan]

Hi,

Does anyone has a better way of solving this problem? It worked before but now it either crashes without giving any warning or in the message panel, it says the sto. file cannot be found. Please help.

I loaded the setting file which I saved before and it worked, and now it crash! How is that possible!!!!!

A million thanks!

Sylvia

[Matthias Kunze]

I think you need an .mot or .sto-file (results from IK) for input in the ID-tool and another .mot or .sto-file for the kinematics of the external loads. The most reason in my runs of the ID-tool with crashes was that all the files wasn't in the same directory or pointed to the wrong path/name of the file.

The names in the file header of the setting files have to be also the same like the real file-name.


Hope this helps,

Matthias