Search Unity

  1. We are migrating the Unity Forums to Unity Discussions. On July 12, the Unity Forums will become read-only. On July 15, Unity Discussions will become read-only until July 18, when the new design and the migrated forum contents will go live. Read our full announcement for more information and let us know if you have any questions.
    Dismiss Notice
  2. Dismiss Notice

What does jointVelocity[] return?

Discussion in 'Robotics' started by Domi47, Sep 27, 2021.

  1. Domi47

    Domi47

    Joined:
    Feb 4, 2021
    Posts:
    10
    Hello everyone,

    I have some doubts about using joinVelocity[] (https://docs.unity3d.com/2020.2/Documentation/ScriptReference/ArticulationBody-jointVelocity.html) as I'm getting some strange values.

    I'm using the position Control as in the documentation (https://github.com/Unity-Technologi...ppendix.md#guide-to-write-your-own-controller). I calculate a DeltaStep to reach a target in a given time:

    Code (CSharp):
    1. distance = targetPos - currentPos;
    2. vel = distance/time_sec;
    3. deltaStep = vel * Time.fixedDeltaTime;
    Then, in a FixedUpdate(), I use this deltaStep to update the joint target and so reach the final one. I also get the joint position and velocity:
    Code (CSharp):
    1. ArticulationDrive joint_xDrive = joint.xDrive;
    2. joint_xDrive.target += deltaStep;
    3. joint.xDrive = joint_xDrive;
    4.  
    5. posMsg.data = joint.jointPosition[0];
    6. velMsg.data = joint.jointVelocity[0];
    When I use it in a revolute joint I set the deltaStep as degrees, and joint.jointPosition[0] returns the current position in radians. But what is joint.jointVelocity[0] returning? For example when trying to get 1.0 radians from the origin at 2.5 seconds, it means a distance of approximately 57.30 deg, and a velocity of 22.92 deg/sec (0.4 rad/sec). With a fixedDeltaTime of 0.02 sec the value of deltaStep will be 0.4584 deg (0.00799 rad). In this case joint.jointVelocity[0] returns a value similar to 0.0091 (what is also similar to my deltaStep in radians). Is it maybe the applied deltaStep in radians?

    But in case of using a Prismatic joint the returned value by joint.jointVelocity[0] is more confusing. I set a deltaStep in meters and joint.jointPosition[0] returns the current position in meters too. But if for example I try to get to 0.015 meters from the origin in 2.5 sec, this means a distance of 0.015m and a velocity of 0.006 m/s. With a fixedDeltaTime of 0.02 I will have a deltaStep of 0.00012 m. In this case when using joint.jointVelocity[0] I get a value of 0.0012, so the order of magnitude is different from my deltaStep. In this case, Is it returning an applied deltaStep in decimeters?

    I hope I explained it well. The version of ROS TCP Connector and URDF Importer that I'm using is 0.4.0 as I started the project with these ones.

    Thank you.
     
  2. yant

    yant

    Unity Technologies

    Joined:
    Jul 24, 2013
    Posts:
    597
    I might need to repro your scene locally, but all of our reduced coordinate space getters/setters work on the data provided by PhysX directly -- that is angles are in rad, distances are in meters, and the velocities are direct time derivatives of those.
     
  3. yant

    yant

    Unity Technologies

    Joined:
    Jul 24, 2013
    Posts:
    597
    Additionally, I wanted to point out that currently we have an asymmetry in the API:
    * ArticulationBody's angular drive parameters are always in degrees (e.g. target & targetVelocity)
    * When in reduced space, we use the raw values from the physics engine -- so angles come out as radians.

    Could this be the source of the problem you're observing?
     
  4. Domi47

    Domi47

    Joined:
    Feb 4, 2021
    Posts:
    10
    Thank you for your answers @yant . Yes I realized about that asymmetry you mentioned, that's why I wanted to confirm what it is returning. So for a revolute joint it should return rad/s and for a prismatic joint it uses m/s right?
    Then I'm still having an error somewhere cause as I said in my example when I try to reach 0.4 rad/s (using a deltaStep of 0.4584 deg and a FixedDeltaTime of 0.02), it returns something like 0.0091 .