Krishna/feat/openarm damiao

[KrishnaH96]

Contribution path

• Linked issue or discussion: feat(manipulation): add OpenYAM Damiao adapter #3129 (this PR is stacked on that branch and extends its pattern to OpenArm)

Problem

OpenArm support predates the whole-body Damiao stack: a hand-rolled CAN driver plus a manipulator-protocol adapter, one instance per bus, grippers present on the bus but never wired, no sim/real switch, and robot-specific gravity compensation code.

Solution

Stacked on #3129. This ports OpenArm to the generic DamiaoWholeBodyAdapter, mirroring the OpenYAM integration:

• OpenArmDamiaoAdapter: both v10 arms (2x DM8006, 2x DM4340, 3x DM4310, send ids 0x01..0x07) and both grippers (DM4310 at 0x08) as one whole-body device over two CAN buses (left=can1, right=can0), commanded in one synchronized tick. Gravity compensation uses the bimanual URDF from openarm_description (validated in pinocchio: 14 joints, order left1..7 then right1..7, finite torques).
• Config rewritten on the OpenYAM pattern: one 16-joint WHOLE_BODY component, mock/real selection via global_config.simulation, hardware-measured MIT gains carried over from the legacy adapter, and per-side planning models now declare an explicit coordinator-to-URDF joint name mapping.
• Blueprints: coordinator-openarm, openarm-planner-coordinator, keyboard-teleop-openarm, and keyboard-teleop-openarm-planner. The keyboard jogs the left arm while the right arm's twist task holds pose. The [ and ] keys drive both grippers through a single servo task, enabled by a new KeyboardTeleopConfig.gripper_joint_names field.
• The legacy driver, adapter, and CAN scripts are removed as superseded. Registry golden sets are updated in both directions. The e2e planning-groups test moves to openarm-planner-coordinator, since the test harness passes --simulation and now gets the in-memory adapter automatically.

How to Test

dimos can setup can0
dimos can setup can1
dimos run keyboard-teleop-openarm

Automated validation:

uv run pytest dimos/hardware/whole_body dimos/hardware/test_adapter_registries.py
uv run pytest dimos/robot/test_all_blueprints.py -k openarm

Not validated on hardware (review focus welcome)

• Gripper opening_direction and default_current were copied from the OpenYAM values and need bench verification.
• Bus-side mapping (can0/can1 comes from USB enumeration order) is unverified on the rig.
• Motor models are v1.0 facts taken from the legacy driver. OpenArm 2.0 units may ship DM8009 shoulders; we will re-probe before first enable.
• Open design question: there is no two-bus deployment override in global config. The scalar global_config.can_port is deliberately not wired; overrides go through DamiaoRuntimeConfig.bus_addresses.

AI assistance

Claude Code with Fable 5 assisted throughout: studying the #3129 adapter stack, implementing the adapter, config, blueprints, and tests, validating the bimanual gravity URDF in pinocchio, and writing the docs. The author directed the design decisions (bimanual-first scope, gain carryover from the legacy adapter, legacy driver removal) and reviewed the changes. Mock and unit test validation only; physical OpenArm motion will be tested and findings willl be updated here.

Checklist

• I have read and approved the CLA.

[TomCC7]

lgtm overall. however, there's some other changes needed to make sure everything is working with openarm:

1. manipulation module will definitely fail because current srdf generation does not support dual arm setup (on intention). We need to define the srdf and feed manually
2. how to support multiple gripper, but this can be a separate pr

And a general (safe) testing tip: don't write the command to robot (can just comment out the write function call) and just open visualization to make sure robot joint looks correct (read working), and once it's verified we can enable motor write

[codecov]

❌ 1 Tests Failed:

Tests completed	Failed	Passed	Skipped
3930	1	3929	68

View the top 1 failed test(s) by shortest run time

dimos.e2e_tests.test_manipulation_planning_groups::test_dual_arm_plans_and_dispatches_both_arms_through_control_coordinator

Stack Traces | 18.5s run time

dimos.protocol.rpc.rpc_utils.RemoteError: [Remote builtins.RuntimeError] Invalid goal configuration requested, cannot plan!

Remote traceback:
Traceback (most recent call last):
  File ".../protocol/rpc/pubsubrpc.py", line 280, in execute_and_respond
    response = f(*args[0], **args[1])
               ^^^^^^^^^^^^^^^^^^^^^^
  File ".../protocol/rpc/spec.py", line 116, in override_f
    return getattr(module, fname)(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File ".../dimos/manipulation/manipulation_module.py", line 1044, in plan_to_joint_targets
    return self.generate_plan_to_joint_targets(joint_targets) is not None
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File ".../dimos/manipulation/manipulation_module.py", line 1083, in generate_plan_to_joint_targets
    return self._plan_selected_path(group_ids, start, goal, planning_epoch)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File ".../dimos/manipulation/manipulation_module.py", line 773, in _plan_selected_path
    result = self._planner.plan_selected_joint_path(
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File ".../planning/world/roboplan_world.py", line 523, in plan_selected_joint_path
    return self._plan_group(
           ^^^^^^^^^^^^^^^^^
  File ".../planning/world/roboplan_world.py", line 1134, in _plan_group
    result = self._run_native_rrt(
             ^^^^^^^^^^^^^^^^^^^^^
  File ".../planning/world/roboplan_world.py", line 1178, in _run_native_rrt
    result = planner.plan(start, goal)
             ^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: Invalid goal configuration requested, cannot plan!


The above exception was the direct cause of the following exception:

lcm_spy = <dimos.e2e_tests.lcm_spy.LcmSpy object at 0x7b70222f6750>
start_blueprint = <function start_blueprint.<locals>.set_name_and_start at 0x7b7021ba7920>

    def test_dual_arm_plans_and_dispatches_both_arms_through_control_coordinator(
        lcm_spy: LcmSpy,
        start_blueprint: Callable[..., DimosCliCall],
    ) -> None:
        """Plan one generated plan over both arms and dispatch through one trajectory task."""
        _start_openarm_mock_planner(start_blueprint, lcm_spy)
    
        client = RPCClient(None, ManipulationModule)
        coordinator_client = RPCClient(None, ControlCoordinator)
        try:
            info = _wait_for_robot_info(client, ROBOT_NAME)
            group_ids = _planning_group_ids(info)
    
            tasks = coordinator_client.list_tasks()
            assert tasks == [DEFAULT_TRAJECTORY_TASK_NAME]
    
            _prepare_for_planning(client, (ROBOT_NAME,))
    
>           planned = client.plan_to_joint_targets(
                {
                    group_ids["left_manipulator"]: _offset_target(client, LEFT_SLICE, 0.02),
                    group_ids["right_manipulator"]: _offset_target(client, RIGHT_SLICE, -0.02),
                }
            )

client     = <dimos.core.rpc_client.RPCClient object at 0x7b702323e360>
coordinator_client = <dimos.core.rpc_client.RPCClient object at 0x7b70222f6960>
group_ids  = {'left_manipulator': 'openarm/left_manipulator', 'right_manipulator': 'openarm/right_manipulator'}
info       = {'base_link': 'openarm_body_link0', 'end_effector_link': None, 'has_joint_name_mapping': True, 'home_joints': [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...], ...}
lcm_spy    = <dimos.e2e_tests.lcm_spy.LcmSpy object at 0x7b70222f6750>
start_blueprint = <function start_blueprint.<locals>.set_name_and_start at 0x7b7021ba7920>
tasks      = ['traj_arm']

dimos/e2e_tests/test_manipulation_planning_groups.py:211: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
dimos/core/rpc_client.py:93: in __call__
    result, unsub_fn = self._rpc.call_sync(
        args       = ({'openarm/left_manipulator': JointState(ts=1785900657.7928898, frame_id='', name=[], position=[0.02, 0.02, 0.02, 0.02...7.7935464, frame_id='', name=[], position=[-0.02, -0.02, -0.02, -0.02, -0.02, -0.02, -0.02], velocity=[], effort=[])},)
        kwargs     = {}
        self       = <dimos.core.rpc_client.RpcCall object at 0x7b70222d8bc0>
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

self = <dimos.protocol.rpc.pubsubrpc.LCMRPC object at 0x7b70222857f0>
name = 'ManipulationModule/plan_to_joint_targets'
arguments = (({'openarm/left_manipulator': JointState(ts=1785900657.7928898, frame_id='', name=[], position=[0.02, 0.02, 0.02, 0.0...5464, frame_id='', name=[], position=[-0.02, -0.02, -0.02, -0.02, -0.02, -0.02, -0.02], velocity=[], effort=[])},), {})
rpc_timeout = 120.0

    def call_sync(
        self, name: str, arguments: Args, rpc_timeout: float | None = None
    ) -> tuple[Any, Callable[[], None]]:
        if rpc_timeout is None:
            method = name.rsplit("/", 1)[-1]
            rpc_timeout = self.rpc_timeouts.get(name) or self.rpc_timeouts.get(
                method, self.default_rpc_timeout
            )
        event = threading.Event()
    
        def receive_value(val) -> None:  # type: ignore[no-untyped-def]
            event.result = val  # type: ignore[attr-defined]  # attach to event
            event.set()
    
        unsub_fn = self.call(name, arguments, receive_value)
        if not event.wait(rpc_timeout):
            # Retries register new callbacks. Remove this expired callback so
            # repeated timeouts do not accumulate entries in the shared response map.
            unsub_fn()
            raise TimeoutError(f"RPC call to '{name}' timed out after {rpc_timeout} seconds")
    
        # Check if the result is an exception and raise it
        result = event.result  # type: ignore[attr-defined]
        if isinstance(result, BaseException):
>           raise result
E           RuntimeError: Invalid goal configuration requested, cannot plan!

arguments  = (({'openarm/left_manipulator': JointState(ts=1785900657.7928898, frame_id='', name=[], position=[0.02, 0.02, 0.02, 0.0...5464, frame_id='', name=[], position=[-0.02, -0.02, -0.02, -0.02, -0.02, -0.02, -0.02], velocity=[], effort=[])},), {})
event      = <threading.Event at 0x7b70222d8ad0: set>
method     = 'plan_to_joint_targets'
name       = 'ManipulationModule/plan_to_joint_targets'
receive_value = <function RPCClient.call_sync.<locals>.receive_value at 0x7b7021ba76a0>
result     = RuntimeError('Invalid goal configuration requested, cannot plan!')
rpc_timeout = 120.0
self       = <dimos.protocol.rpc.pubsubrpc.LCMRPC object at 0x7b70222857f0>
unsub_fn   = <function PubSubRPCMixin.call_cb.<locals>.unsubscribe_callback at 0x7b7021bf0a40>

.../protocol/rpc/spec.py:88: RuntimeError

To view more test analytics, go to the Test Analytics Dashboard
_{📋 Got 3 mins? Take this short survey to help us improve Test Analytics.}

[TomCC7]

is this needed? lgtm otherwise