OpenRAVE Documentation

inversekinematics Module

Manages compiled inverse kinematics files for robots using ikfast.



Running: --database inversekinematics


First set the active manipulator, and then instantiate the InverseKinematicsModel class specifying the iktype and free indices.

ikmodel = openravepy.databases.inversekinematics.InverseKinematicsModel(robot,iktype=IkParameterizationType.Transform6D)
if not ikmodel.load():

The supported types are defined by IkParameterizationType and are propagated throughout the entire OpenRAVE framework. All solve methods take in a IkParameterization structure, which handles each IK type’s serialization, distances metrics, derivatives, and transformation.

To show the manipulator and IK results do:

../../_images/inversekinematics_pr2show.jpg --database inversekinematics --robot=robots/pr2-beta-static.zae --manipname=leftarm --show

It is also possible to test the IK on a scene: --database inversekinematics --robot=data/pr2test1.env.xml --manipname=leftarm --show


This database generator uses IKFast: The Robot Kinematics Compiler to generate optimized and stable analytic inverse kinematics solvers for any robot manipulator. The manipulator’s arm joints are used for obtaining the joints to solve for. The user can specify the IK type (Rotation, Translation, Full 6D, Ray 4D, etc), the free joints of the kinematics, and the precision. For example, generating the right arm 6D IK for the PR2 robot where the free joint is the first joint and the free increment is 0.01 radians is: --database inversekinematics --robot=robots/pr2-beta-static.zae --manipname=rightarm  --freejoint=r_shoulder_pan_joint --freeinc=0.01

Generating the 3d rotation IK for the stage below is: --database inversekinematics --robot=robots/rotation_stage.robot.xml --iktype=rotation3d

Generating the ray inverse kinematics for the 4 degrees of freedom barrett wam is: --database inversekinematics --robot=robots/barrettwam4.robot.xml --iktype=ray4d --database inversekinematics --robot=robots/pr2-beta-static.zae --iktype=ray4d --manipname=rightarm_camera --database inversekinematics --robot=robots/neuronics-katana.zae --iktype=translationdirection5d --manipname=arm

The filename that the code is saved in can be retrieved by --database inversekinematics --robot=robots/neuronics-katana.zae --iktype=translationdirection5d --manipname=arm --getfilename


Every IK solver should be tested with the robot using --iktests=XXX. However, calling inversekinematics will always re-generate the IK, even if one already exists. In order to just run tests, it is possible to specify the --usecached option to prevent re-generation and specifically test: --database inversekinematics --robot=robots/barrettwam.robot.xml --usecached --iktests=100

This will give the success rate along with information whether the IK gives a wrong results or fails to find a solution.

If there are a lot of free joints in the IK solver, then their discretization can greatly affect whether solutions are found or not. In this case, it is advisable to reduce the discretization threshold by using the --freeinc option.

Loading from C++

It is possible to use the auto-generation process through c++ by loading the IKFast problem and calling LoadIKFastSolver command.

ikfastloader.cpp - example for loading IK in C++.



Usage: --database inversekinematics [options]

Uses ikfast to compute the closed-form inverse kinematics equations of a robot manipulator, generates a C++ file, and compiles this file
into a shared object which can then be loaded by OpenRAVE.

  -h, --help            show this help message and exit
                        Optional joint name specifying a free parameter of the manipulator. The value of a free joint is known at runtime,
                        but not known at IK generation time. If nothing specified, assumes all joints not solving for are free parameters.
                        Can be specified multiple times for multiple free parameters.
                        The precision to compute the inverse kinematics in, (default=8).
  --usecached           If set, will always try to use the cached ik c++ file, instead of generating a new one.
  --freeinc=FREEINC     The discretization value of freejoints.
  --numiktests=IKTESTS, --iktests=IKTESTS
                        Will test the ik solver and return the success rate. IKTESTS can be an integer to specify number of random tests, it
                        can also be a filename to specify the joint values of the manipulator to test. The formst of the filename is
                        #numiktests [dof values]*
                        Number of IK calls for measuring the internal ikfast solver.
                        If specified, will output the generated code in that language (ie --outputlang=cpp).
  -i, --ipython         if true will drop into the ipython interpreter right before ikfast is called
  --iktype=IKTYPE       The ik type to build the solver current types are: Transform6D, Rotation3D, Translation3D, Direction3D, Ray4D,
                        Lookat3D, TranslationDirection5D, TranslationXY2D, TranslationXYOrientation3D, TranslationLocalGlobal6D,
                        TranslationXAxisAngle4D, TranslationYAxisAngle4D, TranslationZAxisAngle4D, TranslationXAxisAngleZNorm4D,
                        TranslationYAxisAngleXNorm4D, TranslationZAxisAngleYNorm4D, CustomDataBit

  OpenRAVE Environment Options:
                        List all plugins and the interfaces they provide.
                        Default collision checker to use
    --physics=_PHYSICS  physics engine to use (default=none)
    --viewer=_VIEWER    viewer to use (default=qtcoin)
    --server=_SERVER    server to use (default=None).
                        port to load server on (default=4765).
    --module=_MODULES   module to load, can specify multiple modules. Two arguments are required: "name" "args".
    -l _LEVEL, --level=_LEVEL, --log_level=_LEVEL
                        Debug level, one of (fatal,error,warn,info,debug,verbose,verifyplans)
    --testmode          if set, will run the program in a finite amount of time and spend computation time validating results. Used for

  OpenRAVE Database Generator General Options:
    --show              Graphically shows the built model
    --getfilename       If set, will return the final database filename where all data is stored
    --gethas            If set, will exit with 0 if datafile is generated and up to date, otherwise will return a 1. This will require
                        loading the model and checking versions, so might be a little slow.
    --robot=ROBOT       OpenRAVE robot to load (default=robots/barrettsegway.robot.xml)
                        number of threads to compute the database with (default=1)
                        The name of the manipulator on the robot to use

Class Definitions

class openravepy.databases.inversekinematics.InverseKinematicsModel(robot, iktype=None, forceikfast=False, freeindices=None, freejoints=None)[source]

Bases: openravepy.databases.DatabaseGenerator

Generates analytical inverse-kinematics solutions, compiles them into a shared object/DLL, and sets the robot’s iksolver. Only generates the models for the robot’s active manipulator. To generate IK models for each manipulator in the robot, mulitple InverseKinematicsModel classes have to be created.

Parameters:forceikfast – if set will always force the ikfast solver
class ArmVisibility(manip, transparency=1)[source]

When ‘entered’ will hide all the non-arm links in order to facilitate visiblity of the gripper

static InverseKinematicsModel.CreateOptionParser()[source]
static InverseKinematicsModel.RunFromParser(Model=None, parser=None, args=None, **kwargs)[source]
InverseKinematicsModel.generate(iktype=None, freejoints=None, freeinc=None, freeindices=None, precision=None, forceikbuild=True, outputlang=None, ipython=False)[source]
InverseKinematicsModel.getDefaultFreeIncrements(freeincrot, freeinctrans)[source]

Returns a list of delta increments appropriate for each free index


Returns a default set of free indices if the robot has more joints than required by the IK. In the futrue, this function will contain heuristics in order to select the best indices candidates.

static InverseKinematicsModel.getcompiler()[source]
InverseKinematicsModel.getsourcefilename(read=False, outputlang='cpp')[source]
InverseKinematicsModel.load(freeinc=None, checkforloaded=True, *args, **kwargs)[source]

Sets the ik solver on the robot.

freeinc is a list of the delta increments of the freejoint values that can override the default values., options=None, forceclosure=True)[source]

Tests the iksolver.

openravepy.databases.inversekinematics.RaveCreateIkSolver((Environment)env, (str)name) → IkSolver :

OPENRAVE_API IkSolverBasePtr RaveCreateIkSolver(EnvironmentBasePtr penv, const std::string & name)

openravepy.databases.inversekinematics.RaveCreateModule((Environment)env, (str)name) → Module :

OPENRAVE_API ModuleBasePtr RaveCreateModule(EnvironmentBasePtr penv, const std::string & name)

openravepy.databases.inversekinematics.RaveDestroy() → None :

OPENRAVE_API void RaveDestroy()

Destroys the entire OpenRAVE state and all loaded environments.

This functions should be always called before program shutdown in order to assure all resources are relased appropriately.

openravepy.databases.inversekinematics.RaveFindDatabaseFile((str)arg1, (bool)arg2) → str :

OPENRAVE_API std::string RaveFindDatabaseFile(const std::string & filename, bool bRead = true )

Searches for a filename in the database and returns a full path/URL to it.

filename -
the relative filename in the database
bRead -
if true will only return a file if it exists. If false, will return the filename of the first valid database directory.
a non-empty string if a file could be found.
openravepy.databases.inversekinematics.openravepyCompilerVersion() → str :

Returns the compiler version that openravepy_int was compiled with*args, **kwargs)[source]

Command-line execution of the example. args specifies a list of the arguments to the script.


Having problems with OpenRAVE?