The sections below outline our suggestions for collecting high-quality experimental data for use for generating simulations of human and animal motion.


OpenSim simulations are generated from human motion and forces that are experimentally collected and used as inputs to Inverse Kinematics, Inverse Dynamics, Static Optimization, and Computed Muscle Control algorithms. As for any model and algorithm, the quality of the input data is essential for high-quality results. Typically collected motion data, markers and forces, should be collected with familiarity on how these would be used to generate simulations. 

When you start collecting experimental data to analyze motion and generate dynamic simulations, we highly recommend that you develop and document lab protocols and standards. By establishing a protocol for marker sets, camera locations, and force plate coordinate frames, data will be more repeatable and it will be easier to pre-process and import data into OpenSim.

Collecting Marker Data


If we know the positions of three points on a rigid body in three-dimensional space, we can completely determine that rigid body's position and orientation. We recommend developing Marker sets to give you the locations of at least three markers on each rigid body segment so that the position and orientation of each body segment of interest can be completely determined. Marker sets should also include additional markers to identify joint center locations (medial and lateral malleoli for ankle joint center and medial and lateral epicondyles for knee joints). As a general rule, you must place enough markers on your subject to scale and track each body segment you will model. 

  • You need at least three non-collinear markers to track the 6 DOF motion (position and orientation) of a body segment.
  • Try to place markers on anatomical locations that will have the least skin and muscle motion.
  • There are several OpenSim model marker sets that you can use and adapt, including Full-body model used for dynamic simulations of running by Hamner et al., 2010:

To read more about markers sets for motion capture, please see the following references:

  • Cappozzo, A., Catani, F., Croce, U.D., Leardini, A., 1995. Position and orientation in space of bones during movement: anatomical frame definition and determination. Clinical Biomechanics (Bristol, Avon) 10, 171–178.
  • Davis R, Ounpuu S, Tyburski D, Gage J. A gait analysis data collection and reduction technique. Hum Mov Sci 1991;10:575–87.

  • Kadaba, M.P., Ramakrishnan, H.K., Wootten, M.E., 1990. Measurement of lower extremity kinematics during level walking. J. Orthop. Res. 8 (3), 383–392.


Take Photos and Video during Data Collection

  • Take lots of photos and video during experiments so that you can verify marker placement and other factors for the data you collect. Get a tripod and record each trial. Some motion capture systems can even sync to a digital camcorder.
  • Take pictures of your subjects in the static pose. These picture are valuable for evaluating the results of the Scale tool and creating Marker Sets.


Measure as many subject specifics as possible, including height, mass, body segment lengths, mass distribution (if DXA is available), and strength (if a Biodex is available). 

  • You can use this data, along with marker positions, to best match the generic model to a specific subject.
  • To improve scaling of the torso and lower extremities, you can calculate joint centers from functional or regression methods and append the joint centers to your static trial data (see Associating Data with a Motion).  
    • Gamage, S.S., Lasenby, J., 2002. New least squares solutions for estimating the average centre of rotation and the axis of rotation. J Biomech 35, 87-93. 

    • Siston, R.A., Delp, S.L. 2006. Evaluation of a new algorithm to determine the hip joint center. J Biomech 39, 125-130.

Collecting External Forces

  • You must measure all external forces applied to your subject in order to model the full dynamics of the system. This includes ground reaction forces and any forces applied by external objects like a brace or harness. 
  • Record or otherwise determine the magnitude, direction, and point of application for all forces applied to the subject. Make sure you also document what coordinate system the forces are measured in so that you can perform any necessary transformations when you prepare your data for import into OpenSim.
  • Calibrate center-of-pressure measurements with marker positions using a calibration "T" to pinpoint where on the force-plate the point load (lowest tip of the "T") is being applied. If the center-of-pressure calculated from the force-plate does not match the "T" location from markers (within marker resolution), you need align the force-plate and marker mo-cap reference frames.
  • Note that generating dynamic simulations of motion are very sensitive to noise in force plate data whether due to building vibration, treadmill vibration (if using instrumented treadmills), or other sources. The baseline noise levels of force plates should be carefully evaluated before pursuing a simulation-based study and filtering of the data may be necessary to generate simulations.

Measuring EMG

  • If possible, you should collect EMG data for as many muscles as possible, particularly the muscles whose function you are most interested in studying.
  • Collecting maximum voluntary contractions or other measurements for normalization can be very useful for comparing estimated activations from simulations to EMG data.
  • If you don't have access to EMG, search the literature for reported values for similar activities.
  • Access to EMG data from experimental motion capture or the literature will allow you to evaluate whether the muscle activations you calculate in a simulation are reasonable.
  • For more information about collecting EMG, please see the following references:
    • A link to the ABC of EMG booklet. This describes many useful ways to collect and filter EMG data. Free download.
    • Hermens, H. J., Freriks, B., Disselhorst-Klug, C. and Rau, G. (2000). Development of recommendations for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol 10, 361-74.