New test for IMU orientation #65
Conversation
There was a problem hiding this comment.
Great work @martinaxgloria!
We have only to figure out how to handle the new dependencies
| axesList.addString("r_shoulder_pitch"); | ||
| axesList.addString("r_shoulder_roll"); | ||
| axesList.addString("r_shoulder_yaw"); | ||
| axesList.addString("r_elbow"); |
There was a problem hiding this comment.
This list can be parametrized in future improvements in order to be able to test sensors in other kinematic chains
| double minLim; | ||
| double maxLim; | ||
|
|
||
| for (int i = kinDynComp.model().getJointIndex("neck_pitch"); i <= kinDynComp.model().getJointIndex("neck_yaw"); i++) |
There was a problem hiding this comment.
Also this can be generalized it making it a parameter, ideally the user has to specify the 3 joints that have to be controlled, let's remind it for future versions of the test
|
The CI is failing due to |
|
|
||
| auto error = (expectedImuSignal * imuSignal.inverse()).log(); | ||
| double err_mean = (std::accumulate(error.begin(), error.end(), 0)) / error.size(); | ||
| ROBOTTESTINGFRAMEWORK_ASSERT_ERROR_IF_FALSE(err_mean < 0.1, "Error > 0.1! Aborting ..."); |
There was a problem hiding this comment.
Similar to @Nicogene comments, in the future, the threshold could be a parameter, as if for some model/sensors are quite accurate, you may want to make sure that they contibute to be accurate.
There was a problem hiding this comment.
Great work! A possible future improvements (that are not blocking the PR or need to be done at all cost, but I want to list them here to avoid forgetting):
- Support specifying multiple sensors
- Support for automatically testing all the imus (we can get the list from the iDynTree's model)
- Support for saving tests data in a file for future inspection
|
I think I integrated all the proposed changes in 03c32b2, except for the parameterization. Please, let me know if I forgot something |
As per title, this PR added a new test that can be run both in simulation environment and on the real robot to test the Euler angles retrieved from the IMU mounted on the RFE.
So far, the test could be run on iCubGenova11 with:
robottestingframework-testrunner --verbose --test plugins/imu.so --param "--robot icub --model /usr/local/src/robot/robotology-superbuild/build/install/share/iCub/robots/iCubGenova11/model.urdf --port /icub/head/inertials --sensor head_imu_0 --frame head_imu_0"For the future, when probably we'll have to run multiple tests at the same time, there is also the skeleton for the test suites.
A side note (imho it's not so important for the time being): if we want to use the fixture parameter, we have to define and install a .sdf file that will be used to load the model on gazebo. This part is commented out right now, and if someone would launch the test in simulation, the gazebo environment should be launched before on another shell.
I will write all the necessary information in the documentation