Unable to get correct force data form contact sensor

[gyeongjun9]

Hi, I'm trying to use contact sensor for the weight measure of water particels (this is the final purpose), and for the basic step, I'm checking if the contact sensor works well.

I wrote my code like right below.

import argparse

from isaaclab.app import AppLauncher

# create argparser
parser = argparse.ArgumentParser(description="Tutorial on spawning prims into the scene.")
parser.add_argument("--num_envs", type=int, default=3, help="Number of environments to spawn.")
# append AppLauncher cli args
AppLauncher.add_app_launcher_args(parser)
# parse the arguments
args_cli = parser.parse_args()
# launch omniverse app
app_launcher = AppLauncher(args_cli)
simulation_app = app_launcher.app

"""Rest everything follows."""

import torch

import isaaclab.sim as sim_utils
from isaaclab.assets import ArticulationCfg, AssetBaseCfg, RigidObjectCfg
from isaaclab.scene import InteractiveScene, InteractiveSceneCfg
from isaaclab.sensors import ContactSensorCfg
from isaaclab.utils import configclass

# Pre-defined configs
from isaaclab_assets.robots.franka import FRANKA_PANDA_CFG # isort: skip

class SceneCfg(InteractiveSceneCfg):

    # ground plane
    ground = AssetBaseCfg(prim_path="/World/GroundPlane", spawn=sim_utils.GroundPlaneCfg())

    # lights
    dome_light = AssetBaseCfg(prim_path="/World/Light", spawn=sim_utils.DomeLightCfg(intensity=3000.0, color=(0.75, 0.75, 0.75)))

    # articulation
    robot: ArticulationCfg = FRANKA_PANDA_CFG.replace(prim_path="{ENV_REGEX_NS}/Robot")
    """ What is the usage of 'robot : ' ? """

    # # water & cup
    # usd_path_1 = "omniverse://localhost/Projects/Water_Env/Cup_Water_Flattened.usd"
    # cfg = sim_utils.UsdFileCfg(usd_path=usd_path_1)
    # cfg.func("/World/Objects/Cup_Water", cfg, translation=(0.0, 0.0, 1.05))

    # cube & contact sensor
    cube_1 = RigidObjectCfg(
        prim_path="{ENV_REGEX_NS}/Cube_1",
        spawn=sim_utils.CuboidCfg(
            size=(0.5, 0.5, 0.1),
            rigid_props=sim_utils.RigidBodyPropertiesCfg(),
            mass_props=sim_utils.MassPropertiesCfg(mass=100.0),
            collision_props=sim_utils.CollisionPropertiesCfg(),
            physics_material=sim_utils.RigidBodyMaterialCfg(static_friction=1.0),
            visual_material=sim_utils.PreviewSurfaceCfg(diffuse_color=(0.0, 1.0, 0.0), metallic=0.2),
            activate_contact_sensors=True
        ),
        init_state=RigidObjectCfg.InitialStateCfg(pos=(0.5, 0.5, 0.05)),
    )

    cube_2 = RigidObjectCfg(
        prim_path="{ENV_REGEX_NS}/Cube_2",
        spawn=sim_utils.CuboidCfg(
            size=(0.2, 0.2, 0.1),
            rigid_props=sim_utils.RigidBodyPropertiesCfg(),
            mass_props=sim_utils.MassPropertiesCfg(mass=50.0),
            collision_props=sim_utils.CollisionPropertiesCfg(),
            physics_material=sim_utils.RigidBodyMaterialCfg(static_friction=1.0),
            visual_material=sim_utils.PreviewSurfaceCfg(diffuse_color=(0.0, 1.0, 0.0), metallic=0.2),
            activate_contact_sensors=True
        ),
        init_state=RigidObjectCfg.InitialStateCfg(pos=(0.5, 0.5, 0.3)),
    )

    contact_forces_1 = ContactSensorCfg(
        prim_path="{ENV_REGEX_NS}/Cube_1",
        update_period=0.0,
        history_length=6,
        debug_vis=True,
        #filter_prim_paths_expr=["{ENV_REGEX_NS}/Cube_2"],
    )


def run_simulator(sim: sim_utils.SimulationContext, scene: InteractiveScene):
    """Run the simulator"""
    # Define simulation stepping (not understood it yet..)
    sim_dt = sim.get_physics_dt()
    sim_time = 0.0
    count = 0

    # Simulate physics
    while simulation_app.is_running():
        # Reset
        if count % 500 == 0:
            # reset counter
            count = 0
            # reset the scene entities
            # root state
            # we offset the root state by the origin since the states are wrriten in simulation world frame
            # if this is not done, then the robots will be spawned at the (0,0,0) of the simulation world
            root_state = scene["robot"].data.default_root_state.clone()
            root_state[:, :3] += scene.env_origins
            scene["robot"].write_root_pose_to_sim(root_state[:, :7])
            scene["robot"].write_root_velocity_to_sim(root_state[:, 7:])
            # set joint positions with some noise
            joint_pos, joint_vel = (
                scene["robot"].data.default_joint_pos.clone(),
                scene["robot"].data.default_joint_vel.clone(),
            )
            joint_pos += torch.rand_like(joint_pos) * 0.1
            scene["robot"].write_joint_state_to_sim(joint_pos, joint_vel)
            # clear internal buffers
            scene.reset()
            print("[INFO]: Resetting robot state...")
        # Apply default actions to the robot
        # -- generate actions/commands
        targets = scene["robot"].data.default_joint_pos
        # -- apply action to the robot
        scene["robot"].set_joint_position_target(targets)
        # -- write data to sim
        scene.write_data_to_sim()
        # perform step
        sim.step()
        # update sim-time
        sim_time += sim_dt
        count += 1
        # update buffers
        scene.update(sim_dt)

        # print information from the sensors
        # # print(scene["contact_forces_1"].get_current_frame())
        # print("Received force matrix of: ", scene["contact_forces_1"].data.force_matrix_w)
        print("Received contact force of: ", scene["contact_forces_1"].data.net_forces_w)
        print("-------------------------------")

        """
        1. just one cube
        Received contact force of:  tensor([[[1.3181e-04, 0.0000e+00, 9.8100e+02]],

        [[1.3181e-04, 0.0000e+00, 9.8100e+02]],

        [[1.3181e-04, 0.0000e+00, 9.8100e+02]]], device='cuda:0')
        -------------------------------
        ->> right weights
        """

        """
        2. two cubes
        Received contact force of:  tensor([[[ 5.4774e-05, -1.6073e-04,  1.1281e+02]],

        [[-6.3875e-05, -3.0173e-05,  1.1282e+02]],

        [[-1.8863e-05, -2.6133e-04,  1.1282e+02]]], device='cuda:0')
        -------------------------------
        why printed like this?? three sensor environments are all the same!! ->> it's actually all the same on z ais

        ->>not right weights
        """


def main():
    """Main function."""

    # Initialize the simulation context
    sim_cfg = sim_utils.SimulationCfg(dt=0.005, device=args_cli.device)
    sim = sim_utils.SimulationContext(sim_cfg)
    # Set main camera
    sim.set_camera_view(eye=[3.5, 3.5, 3.5], target=[0.0, 0.0, 0.0])
    # design scene. scene instantiation
    scene_cfg = SceneCfg(num_envs=args_cli.num_envs, env_spacing=2.0)
    scene = InteractiveScene(scene_cfg)
    # Play the simulator
    sim.reset()
    # Now we are ready!
    print("[INFO]: Setup complete...")
    # Run the simulator
    run_simulator(sim, scene)


if __name__ == "__main__":
    # run the main function
    main()
    # close sim app
    simulation_app.close()

I checked if the contact sensor works well by two options.

First option was just placing one cube which weights 100kg. As a result, I could confirm it is right working cause on the z axis, the force was 100 * 9.81 = 981N

But the problem was the second option. I put the 50kg-weighted-cube on the first cube, and checked the force data. And the result was 1.1281e+02 and it is actually much smaller than 981N and weird result.. I don't know why it happens like this and I want to know why, and the way to correct this situation.

[gyeongjun9]

For more accurate situation explain, I post the picture of the simulation.
And actually I want to embody scale by contact sensor!

[jtigue-bdai]

In the second option did you uncomment the filter_prim_paths_expr=["{ENV_REGEX_NS}/Cube_2"], line in the contact sensor definition?

[gyeongjun9]

Hi, @jtigue-bdai Thank you for the reply!

OPTION 2
Yes, I checked it again right before with the simulation uncommented filter_prim_paths_expr=["{ENV_REGEX_NS}/Cube_2"], and the force sensor result was still wrong. Here's the screenshot of it.

OPTION 1
And for the recheck, I simulated again with filter_prim_paths_expr=["{ENV_REGEX_NS}/Cube_2"] uncommented and also cube2 configuration uncommented. Here's the screenshot of it.

[RandomOakForest]

Following up on this. Is the contact sensor you are using the Isaac Sim sensor or the PhysX sensor? Something is definitely off.

[jtigue-bdai]

These are IsaacLab sensors. So they utilize the omni.phyiscs.tensors RigidContactView to access net_forces_w/force_matrix_w

[RandomOakForest]

Following up, I'll move this to our Discussions section. Let us know if you need further help.

[RandomOakForest]

Following up on this, we are unable to repro this behavior. Have you found anything on your side that may have been causing this issue? Let us know may we be of further help.