Tormach Robot Programming Language

The TRPL is based on Python. All commands interpreted by the TRPL interpreter are Python code, which means you can use any Python code to control the robot. However, the graphical conversational programming utility only supports re-interpretation of a subset of commands as defined here.

Program structure
- Special Python Method Names
- Python Stubs
Linear and Angular Units
Waypoints
Movement
User and Tool Frames
- User Frames
- Tool Frames
Digital I/O
PathPilot Remote
Notifications
Program Flow
Persistent Parameter Storage
Probing
- Glossary of Probe Errors
Interrupts
Calibration
List of Supported and Deprecated Commands

Program structure

A TRPL program is a simple Python file with a single main function containing the program code. The support RPL commands are imported per default; any additional external libraries can be imported if required.

Note that the program interpreter only can track code execution in the main program file. Blocking commands in external libraries might, therefore, lock up the program execution

The main function is executed in a loop. If you want to break this behavior, use the exit statement.

   from robot_command.rpl import *  # import all robot commands
   set_units("mm", "deg")  # set the linear and angular unit types for the program

   # waypoints are defined in the header of the program
   waypoint_1 = j[0.764, 1.64, 0.741, 0.433, 0.140, 2.74]
   waypoint_2 = p[361.53,947.19, 937.11, 45, 0, 0]

   # the main function contains the program code
   def main():
       movel(waypoint_1)  # a linear move command
       sleep(0.5)  # wait for 0.5 seconds
       movej(waypoint_2) # a joint move command
       sleep(0.5)  # wait for 0.5 seconds

Special Python Method Names

The following Python method names have a special meaning in the TRPL interpreter:

main: The robot programs main loop. Required in every program.
on_abort: Called when the program is aborted as a result of an error or controlled exit by the operator.
on_pause: Called when the program is paused.

The on_abort and on_pause methods are useful for disabling digital outputs or cleaning up initialized resources. If an error occurs in the on_pause method on_abort will be called as a fallback. If an error occurs in the on_abort method the program will exit instantly.

Python Stubs

Below you can find Python stub files, supported in most modern IDEs.

To use them, install the robot_command-stubs package by issuing the python3 setup.py install command.

robot_command-stubs.zip

Linear and Angular Units

The TRPL supports multiple linear and angular measurement units. It is advised to define the units used in a program as the first line of the program. Changing units at runtime is supported, but highly discouraged.

Supported linear unit types:

“m” - Meters
“mm” - Millimeters
“in” - Inches

Supported angular unit types:

“deg” - Degrees
“rad” - Radians

Sets the active linear, angular and time units for the program.

Parameters:

linear – Linear/length unit type: m, mm, inch
angular – Angular/rotation unit type: deg, rad
time – Time unit type: s, min, h

Examples

set_units("mm", "deg", "s")
set_units(linear="in")
set_units(angular="rad")
set_units(time="s")

robot_command.rpl.get_units(with_time: bool = False) -> Tuple[str, str] | Tuple[str, str, str]

Returns the active linear, angular and time units from the program.

Parameters:	time – return the time unit type when set to True
Returns:	Linear/length unit type, angular/rotation unit type and time unit type

Examples

linear, angular = get_units()
linear, angular, time = get_units(with_time=True)

Waypoints

Waypoints are defined either as joint configuration, as list of joint positions, or as a robot pose with location in Cartesian coordinates XYZ and orientation in Euler angles ABC. Robot poses furthermore relate to a frame from the world origin, usually located at the robot base. If no frame is active or specified with the waypoint, reference to the world origin is assumed.

Global and Local Waypoints

Global waypoints are waypoints that can be reused between programs. These waypoints are referenced by name inside the TRPL. Contrary, local waypoints are defined in the beginning of a robot program as Python variables.

Joints - Waypoint Defined as Joint Configuration

Joint positions can be defined by using the Joints object or the JointsFactory using the j[] shortcut. The values of j1 to j6 are the six joint positions of the robot arm.

class robot_command.rpl.Joints(j1: float = 0.0, j2: float = 0.0, j3: float = 0.0, j4: float = 0.0, j5: float = 0.0, j6: float = 0.0)

The Joints object consists of six joint position values to be used as target for move commands or to represent the current robot joint state.

Examples

waypoint_1 = Joint(323.62, 345.37, 477.76, 431.10, 918.62)
waypoint_2 = Joint(j3=0.543) # all other joint positions are 0.0 by default

copy() -> Joints

Creates a copy of the joints object.

Returns:

Copy of the joints object.

static from_list(joint_list: List[float]) -> Joints

Creates a new joint object from a list of joint positions.

Parameters:	joint_list – List of the six joint positions.
Returns:	New joints object.

from_ros_units(angular_unit: str | SharedRegistryObject | None = None)

Converts the joints object from native ROS units to the target units, removing the unit type if any.

Parameters:	angular_unit – Unit type for the angular joint positions.
Returns:	The resulting joints object.

to_list() -> List[float]

Convert the joints object to a list of joint positions.

Returns:

List of the six joint positions.

to_ros_units(angular_unit: str | SharedRegistryObject | None = None) -> Joints

Converts the joints object to native ROS units, removing the unit type if any. This is useful if you want to send the resulting data to a ROS service.

Parameters:	angular_unit – Unit type for the angular joint positions.
Returns:	The resulting joints object.

with_units(angular_unit: str | SharedRegistryObject | None = None) -> Joints

Adds a unit type to all positions of the joints object. The defaults are the native ROS units. In case a joint position already has units, the unit type is converted accordingly.

Parameters:	angular_unit – Unit type for the angular joint positions.
Returns:	The resulting joints object.

without_units(angular_unit: str | SharedRegistryObject | None = None) -> Joints

Removes units from the joint positions if any. If no unit type is specified ROS units are assumed.

Parameters:	angular_unit – Unit type for the angular joint positions.
Returns:	The resulting joints object.

class robot_command.rpl.JointsFactory

The JointsFactory class helps constructing Joints object using a shorthand notation. In the robot program it can be accessed using the j[] shortcut or the j() shortcut which supports keyword arguments.

Examples

waypoint_1 = j[0.764, 1.64, 0.741, 0.433, 0.140, 2.74]
waypoint_2 = j(j3=0.543) # all other joint positions are 0.0 by default

robot_command.rpl.get_joint_values() -> Joints

Returns the current joint values.

Returns:

Current joint values.

Examples

joint_value = get_joint_values()

Pose - Waypoint Defined as Pose

Poses can be defined using the Pose object or the PoseFactory using the p[] shortcut. x, y, z are the Cartesian coordinates of the robot position. a, b, c is the orientation defined in Euler angles (static XYZ).

Optionally, poses can be defined with a reference frame defined as frame in the robot UI. When no frame is defined, it is assumed that the waypoint is defined in world coordinates.

User frames are not enforced at runtime. You still need to use a change_user_frame() statement before an frame becomes active.

A robot pose consists of XYZ position ABC orientation parameters.

Optionally, an frame frame can be recorded with a waypoint.

Examples

waypoint_1 = Pose(483.21, 34.21, 21.59, 42.03, 71.14)
waypoint_2 = Pose(a=0.543) # all other coordinate values are 0.0 per default.

__mul__(other: Pose) -> Pose

Use KDL frame multiplication to apply a frame to a pose.

Parameters:	other – Other pose.
Returns:	New pose object.

new_wp = Pose(x=10) * waypoint_1 # translates waypoint_1 by x=10
old_wp = Pose(x=10).inverse() * new_wp # translates new_wp back

copy() -> Pose

Creates a copy of the pose object.

Returns:

A copy of the pose.

static from_kdl_frame(frame: Frame) -> Pose

Converts a KDL frame to a pose object.

Parameters:	frame – KDL frame.
Returns:	New pose object.

static from_list(pose_list: List[float]) -> Pose

Creates a new pose object from a list of coordinates.

Parameters:	pose_list – List of the six coordinates.
Returns:	New pose object.

static from_ros_pose(pose: Pose | PoseStamped) -> Pose

Converts a ROS native pose to a pose object.

Parameters:	pose – The ROS stamped pose.
Returns:	New pose object.

from_ros_units(linear_unit: str | SharedRegistryObject | None = None, angular_unit: str | SharedRegistryObject | None = None)

Converts the pose from native ROS units to the target units, removing the unit type.

Parameters:	linear_unit – Unit type for linear coordinates. angular_unit – Unit type for angular coordinates.
Returns:	The resulting pose.

inverse() -> Pose

Creates the inverse of the pose. Useful for calculating frames.

Returns:

New pose object.

to_kdl_frame() -> Frame

Converts the pose object to a KDL frame.

Returns:

KDL frame.

to_list() -> List[float]: Convert the pose to a list of the coordinates. :return: List of the six coordinates.

to_ros_pose() -> Pose

Converts the pose object to a native ROS pose.

Returns:

ROS pose.

to_ros_units(linear_unit: str | SharedRegistryObject | None = None, angular_unit: str | SharedRegistryObject | None = None) -> Pose

Converts the pose to native ROS units, removing the unit type if any. This is useful if you want to send the resulting data to a ROS service.

Parameters:	linear_unit – Unit type for linear coordinates. angular_unit – Unit type for angular coordinates.
Returns:	The resulting pose.

with_units(linear_unit: str | SharedRegistryObject | None = None, angular_unit: str | SharedRegistryObject | None = None) -> Pose

Adds a unit type to all coordinates of the pose. The defaults are the native ROS units. In case a coordinate already has units, the unit type is converted accordingly.

Parameters:	linear_unit – Unit type for linear coordinates. angular_unit – Unit type for angular coordinates.
Returns:	The resulting pose.

without_units(linear_unit: str | SharedRegistryObject | None = None, angular_unit: str | SharedRegistryObject | None = None) -> Pose

Removes units from the coordinates if any. If no unit type is specified ROS units are assumed.

Parameters:	linear_unit – Unit type for linear coordinates. angular_unit – Unit type for angular coordinates.
Returns:	The resulting pose.

class robot_command.rpl.PoseFactory

The PoseFactory class helps constructing Pose object using a shorthand notation. In the robot program it can be accessed using the p[] shortcut or the p() shortcut which supports keyword arguments.

Examples

waypoint_1 = p[202.73, 750.08, 91.75, 6.63, 53.21, "table"]  # captured with table user frame
waypoint_2 = p(y=10.0, frame="table")  # all other coordinate values are 0.0 per default.

robot_command.rpl.get_pose(apply_user_frame: bool = True, apply_tool_frame: bool = True) -> Pose

Returns the current robot pose.

Parameters:	apply_user_frame – Applies the active user frame to the world pose. apply_tool_frame – Applies the active tool frame to the world pose.
Returns:	Current robot pose.

Examples

current_pose = get_pose()

robot_command.rpl.to_local_pose(global_pose, apply_work_offset: bool = True, apply_tool_offset: bool = True) -> Pose

Converts a global pose to a local pose

Parameters:	global_pose – Global workspace Pose to convert to local coordinates (based on specified arguments) apply_work_offset – Applies the active work offset apply_tool_offset – Applies the active tool offset
Returns:	converted local pose

Examples

local_pose = to_local_pose(global_pose)
work_only_pose = to_local_pose(global_pose, apply_tool_offset=False)

Setting and getting global waypoints from robot programs

Robot programs can set or get a global waypoint as part of its execution. Setting of a new global waypoint is achieved by calling robot_command.rpl.set_global_waypoint(). Getting of existing global waypoints is accomplished by calling robot_command.rpl.get_global_waypoint().

robot_command.rpl.set_global_waypoint: alias of funct

robot_command.rpl.get_global_waypoint: alias of funct

Movement

The robot can be moved using the following move types:

movej - Joint Move

robot_command.rpl.movej(target: Pose | Joints, v: float = None, probe: int = 0, velocity_scale: float = 1.0) -> Tuple[int, Time, Joints, Pose] | None

Moves the robot end effector to the target waypoint with a joints move. Targets can be local waypoints or global waypoints defined as pose or joints.

Parameters:

target – target waypoint or joints target
velocity_scale – scale factor for velocity (default is full speed)
probe – specify the probe mode (2-6, or 0 for no probing) Probe mode 2: look for rising edge on probe signal (i.e. contact), raise ProbeFailedError if move completes without seeing a rising edge Probe mode 3: like mode 2 but does not raise error if move completes without rising edge Probe mode 4: like mode 2 but looks for falling edge Probe mode 5: like mode 4 but does not raise an error if move completes without falling edge Probe mode 6: “retract” mode, ignore falling edges and allow motion while probe signal is active, but raise ProbeUnexpectedContactError if a rising edge is seen
v –
scale factor for velocity (default is full speed)

Deprecated since version 3.1.1: use velocity_scale instead

Returns:

tuple of probe results (for probing mode 2,3,4,5) or None: (probe contact type (0 = no contact, 1 = rising, 2 = falling), time of probe contact, Joint positions at probe contact, End-effector position / orientation pose at probe contact)

Examples

movej(waypoint_1)
movej("global_waypoint_1", velocity_scale=0.6)
movej(p[0, 100, 0, 90, 20, 0])

movel - Linear Move

Moves the robot end effector in a straight line from the current position to the target waypoint. Targets can be local waypoints or global waypoints defined as pose or joints.

Parameters:

target – target waypoint
probe – specify the probe mode (2-6, or 0 for no probing) Probe mode 2: look for rising edge on probe signal (i.e. contact), raise ProbeFailedError if move completes without seeing a rising edge Probe mode 3: like mode 2 but does not raise error if move completes without rising edge Probe mode 4: like mode 2 but looks for falling edge Probe mode 5: like mode 4 but does not raise an error if move completes without falling edge Probe mode 6: “retract” mode, ignore falling edges and allow motion while probe signal is active, but raise ProbeUnexpectedContactError if a rising edge is seen
velocity – move velocity as absolute value, interpreted in terms of currently set machine units if quantity without units is given.
accel – move acceleration as absolute value, interpreted in terms of currently set machine units if quantity without units is given.
accel_scale – move acceleration scaling factor 0.0 - 1.0
duration – target move duration in seconds. If move duration based on other inputs is longer, the planned duration will be used.
strict_limits – Enforces strict limits. Moves violating the velocity and acceleration limits will error.
v –
move velocity scaling factor 0.0 - 1.0

Deprecated since version 3.1.1: use velocity instead
a –
move acceleration scaling factor 0.0 - 1.0

Deprecated since version 3.1.1: use accel_scale instead

Returns:

tuple of probe results: (probe contact type (0 = no contact, 1 = rising, 2 = falling), time of probe contact, Joint positions at probe contact, End-effector position / orientation pose at probe contact)

Examples

movel(waypoint_1)
movel("global_waypoint_1", velocity=100)
movel(j[0.764, 1.64, 0.741, 0.433, 0.140, 2.74])

movec - Circular Move

Circular/Arc move command.

Parameters:

interim – interim waypoint
target – target waypoint
probe – specify the probe mode (2-6, or 0 for no probing) Probe mode 2: look for rising edge on probe signal (i.e. contact), raise ProbeFailedError if move completes without seeing a rising edge Probe mode 3: like mode 2 but does not raise error if move completes without rising edge Probe mode 4: like mode 2 but looks for falling edge Probe mode 5: like mode 4 but does not raise an error if move completes without falling edge Probe mode 6: “retract” mode, ignore falling edges and allow motion while probe signal is active, but raise ProbeUnexpectedContactError if a rising edge is seen
velocity – move velocity as absolute value, interpreted in terms of currently set machine units if quantity without units is given.
accel – move acceleration as absolute value, interpreted in terms of currently set machine units if quantity without units is given.
accel_scale – move acceleration scaling factor 0.0 - 1.0
duration – target move duration in seconds. If move duration based on other inputs is longer, the planned duration will be used.
strict_limits – Enforces strict limits. Moves violating the velocity and acceleration limits will error.
v –
move velocity scaling factor 0.0 - 1.0

Deprecated since version 3.1.1: use velocity instead
a –
move acceleration scaling factor 0.0 - 1.0

Deprecated since version 3.1.1: use accel_scale instead

Returns:

tuple of probe results (for probing mode 2,3,4,5) or None: (probe contact type (0 = no contact, 1 = rising, 2 = falling), time of probe contact, Joint positions at probe contact, End-effector position / orientation pose at probe contact)

Examples

movec(waypoint_1, waypoint_2)

movef - Free-form Move

robot_command.rpl.movef(target: Pose | Joints) -> None

Free move command.

Parameters:

target – target target

Trajectory Execution

In same cases it is beneficial to execute a raw trajectory pre-planned in another program, for example Maya Mimic. The robot program supports loading, saving and executing such trajectories.

robot_command.rpl.load_trajectory(file_path: str) -> JointTrajectory

Loads a raw joint trajectory from a CSV file and returns them in ROS format.

Parameters:	file_path – Path of the CSV file.
Returns:	A ROS joint trajectory.

robot_command.rpl.save_trajectory(file_path: str, trajectory: JointTrajectory) -> None

The save trajectory command saves a joint trajectory to a CSV file.

Parameters:

file_path – Path of the CSV file.
trajectory – The joint trajectory to save.

robot_command.rpl.execute_trajectory(trajectory: JointTrajectory, v: float = None, retime: bool = False, velocity_scale: float = 1.0) -> None

The execute trajectory command executes a ROS JointTrajectory. Before running the joint trajectory, the robot moves to the start joint position using a joint move command. :param velocity_scale: move velocity scaling factor 0.0 - 1.0 :param trajectory: a ROS joint trajectory :param retime: Enable retiming the trajectory to make use of the velocity scaling.

Parameters:

v –

move velocity scaling factor 0.0 - 1.0

Deprecated since version 3.1.1: use velocity_scale instead

Examples

trajectory = load_trajectory('test.csv')
execute_trajectory(trajectory)
save_trajectory('test2.csv', trajectory)

Glossary of Move Errors

Move commands typically fails due to two major reasons, either during planning or during execution. You can catch those errors and retry a move inside the robot program.

exception robot_command.rpl.MovePlanningError

exception robot_command.rpl.MoveExecutionError(message=None, error_code=None)

Examples

try:
    movel(waypoint)
except MovePlanningError:  # is raised in case of a planning error
    notify("Move planning failed", warning=True)

Path Blending

Joint moves as well as linear and circular moves can be blended together into one continuous motion. This behavior can be enabled using the set_path_blending() method. Note that the motion will be executed before the next non-motion command. To force the execution of blended commands, use the the sync() method.

robot_command.rpl.set_path_blending(enable: bool, blend_radius: float | None = None) -> None

Enables or disables path blending and sets the blend radius.

Parameters:

enable – Enable or disable path blending.
blend_radius – The blend radius between moves in meters.

Examples

set_path_blending(True, 0.0)  # enable path blending, blend radius 0.0m
movej(waypoint1)
movej(waypoint2)
movej(waypoint3)
sync()  # moves executed before this command
set_path_blending(False)  # disable path blending again

robot_command.rpl.sync() -> None: The sync command is used in wait cycles and to force the execution of queued move commands.

User and Tool Frames

The TRPL supports an arbitrary number of named user and tool frames. These frames are usually defined in the robot UI, however, can also be set inside a program.

User Frames

robot_command.rpl.change_user_frame(name: str | None) -> None

Change the currently active user frame. If an empty string or None is used as the name parameter, the empty user frame world becomes active.

Parameters:

name – The name of the tool frame to activate or None to disable user frames.

Examples

change_user_frame("table")
change_user_frame(None) # disable any active frames

robot_command.rpl.set_user_frame(name: str, pose: Pose | str | None = None, position: Pose | str | None = None, orientation: Pose | str | None = None) -> None

Sets a user frame using a pose, position or orientation or clears an frame.

The position and orientation arguments can be combined to overwrite the pose’s position or orientation.

Parameters:

name – Name of the user frame
pose – Pose to use for the user frame
position – Use the position of this pose to override the position of the pose.
orientation – Use the orientation of this pose to override the orientation of the pose.

Examples

set_user_frame("table", p[0, 100, 0, 0, 0, 0])
set_user_frame("frame_1", waypoint_2, orientation=Pose(a=90))
set_user_frame("frame_1") # clears frame_1

robot_command.rpl.get_user_frame(name: str) -> Pose | None

Returns the pose of a user frame.

Parameters:	name – Name of the user frame.
Returns:	Pose of the user frame.
Raises:	TypeError – if no user frame with the name is found

Examples

pose = get_user_frame("table")

robot_command.rpl.user_frame(pose=None, position=None, orientation=None, world=False)

Scoped frame command. Applies a user frame temporarily on top of the currently active user frame.

The scoped frame command can be used to automatically switch the active frame back to a previous state when the scope is left. Scoped frames can be nested. Scoped frames are temporary and do not have a name.

Parameters:

pose – The frame pose.
position – A pose from which the position is used for the frame.
orientation – A pose from which the orientation is used for the frame.
world – If set to True the frame is absolute.

Examples

with user_frame(p[0, 100, 0, 90, 20, 0]): # creates a temporary frame and activates it
    movel(Pose(x=10)) # move x by 10 starting from the frame
    # the active frame automatically reset when the scope is left

robot_command.rpl.get_active_user_frame() -> str

Returns the name of the active user frame.

Examples

active_user_frame_name = get_active_user_frame()

Tool Frames

robot_command.rpl.change_tool_frame(name: str | None) -> None

Change the currently active tool frame. If an empty string or None is used as the name parameter, the empty tool frame none becomes active.

Parameters:

name – The name of the tool frame to activate or None to disable tool frames.

Examples

change_tool_frame("table")
change_tool_frame(None) # disable any active frames

robot_command.rpl.set_tool_frame(name: str, pose: Pose | str = None, position: Pose | str = None, orientation: Pose | str = None) -> None

Sets a tool frame using a pose, position or orientation or clears an frame.

The position and orientation arguments can be combined to overwrite the pose’s position or orientation.

Parameters:

name – Name of the tool frame
pose – Pose to use for the tool frame
position – Use the position of this pose to override the position of the pose.
orientation – Use the orientation of this pose to override the orientation of the pose.

Examples

set_tool_frame("some_tool", p[0, 0, 100, 0, 0, 0])
set_tool_frame("other_tool", waypoint_2, position=Pose(z=0.1))
set_tool_frame("other_tool") # clears frame other_tool

robot_command.rpl.get_tool_frame(name: str) -> Pose | None

Returns the pose of a tool frame.

Parameters:	name – Name of the tool frame.
Returns:	Pose of the user frame or None if it does not exist.
Raises:	TypeError – if no tool frame with the name is found

Examples

pose = get_tool_frame("tool1")

robot_command.rpl.get_active_tool_frame() -> str

Returns the name of the active tool frame.

Examples

active_tool_frame_name = get_active_tool_frame()

Digital I/O

Digital input/output pins can be accessed via their name or by their number.

robot_command.rpl.set_digital_out(nr_or_name: int | str, state: bool) -> None

Sets a digital output pin to high or low state.

Parameters:

nr_or_name – The number or name of the digital output pin.
state – Set to True or False for on and off.

Examples

set_digital_out("gripper", True)
set_digital_out(2, False)

robot_command.rpl.get_digital_in(nr_or_name: str | int) -> bool

Returns the current digital input state.

Parameters:	nr_or_name – The number or name of the digital output pin.
Returns:	True or False for High and Low.

Examples

io_state = get_digital_in("gripper")
x = get_digital_in(3)

PathPilot Remote

The PathPilot remote interface can be used to remotely control a PathPilot instance.

robot_command.rpl.pathpilot_cycle_start(instance: str = '') -> None

Starts a cycle on the remote PathPilot instance. If no instance argument is given, the command is executed on the first connected PathPilot instance.

Parameters:

instance – PathPilot instance on which cycle start should be executed.

Examples

pathpilot_cycle_start()
pathpilot_cycle_start("left_mill")

robot_command.rpl.pathpilot_mdi(command: str, instance: str = '') -> None

Starts an MDI command on the remote PathPilot instance. If no instance argument is given, the command is executed on the first connected PathPilot instance.

Parameters:

command – MDI command to execute.
instance – PathPilot instance on which this command shall be executed.

Examples

pathpilot_mdi("G0 X10")
pathpilot_mdi("G1 Y-5 F300", "right_mill")

robot_command.rpl.get_pathpilot_state(instance: str = '') -> str

Returns the current state of a PathPilot instance. If no instance argument is given, the command is executed on the first connected PathPilot instance.

Possible states are:

“disconnected” - Instance disconnected
“estop” - Emergency stop active
“running” - a program is running
“ready” - instance is ready to start program
“idle” - instance is idle, no program loaded

Parameters:	instance – PathPilot instance on which cycle start should be executed.
Returns:	The current PathPilot state.

Examples

state = get_pathpilot_instance()
while get_pathpilot_instance("left_mill") != "ready":
    sleep(0.1)

robot_command.rpl.set_machine_frame(pose: Pose | str, instance: str = '') -> None

Sets the origin frame for the 3D visualization of the PathPilot remote machine model.

Parameters:

pose – Pose to use for the machine frame.
instance – Optional machine instance name. If not given, default instance is used.

Examples

set_machine_frame(p[0,0,0,90,0,0], "instance")
set_machine_frame(Pose(x=100))  # sets frame for default instance

Notifications

The TRPL supports interactive user notifications displayed in the robot UI.

robot_command.rpl.notify(message: str, warning: bool = False, error: bool = False, image_path: str = '', timeout: float | None = None) -> None

Creates a popup notification message on the robot UI.

The message argument text is shown to the user.

By default, the message is displayed as informational and thus will not block the program flow. The warning argument shows a warning message, which breaks program flow and can be declined by the operator. The error argument shows a blocking error message, which aborts the program.

The optional image_path argument can be used to display an informational image along with the message

Parameters:

message – Message text to display in the popup.
warning – Set to true if this message is a warning.
error – Set to true if this message is an error message.
image_path – Optional path to an image file to displayed in the popup.
timeout – Optional timeout in seconds.

Examples

notify("Hello World!")
notify("No part found, check the palette.", warning=True)
notify("This should not happen.", error=True, image_path="./fatal_error.png")

robot_command.rpl.input(message: str, default: str = '', image_path: str = '') -> float

Creates a popup input dialog on the robot UI.

The message argument text is shown to the user.

The option default argument can be used to set the default input text.

The optional image_path argument can be used to display an informational image along with the message.

Parameters:	message – Message text to display in the popup. default – The default input value. image_path – Optional path to an image file to displayed in the popup.
Returns:	User input text or None if cancelled

Examples

user_input = input("How many parts should be made?", default="5")
n = int(user_input)

Program Flow

robot_command.rpl.sleep(secs: float) -> None

The sleep command pauses the program execution for t seconds.

Parameters:

secs – sleep time in seconds

Examples

sleep(5.0)

robot_command.rpl.pause(optional: bool = False, active: bool = False) -> None

The pause command pauses the program execution. Equivalent to M01 break.

The optional states whether the pause is optional or not. Optional pause can be enabled in the robot UI by the operator.

Parameters:

optional – If this pause is optional or not.
active – When set to true, this indicates an active pause allowing to operator to jog the program.

Examples

pause()
pause(optional=True)

exit()

The Python builtin exit command instantly aborts the program execution.

Examples

exit()

Loops

The TRPL supports all Python statements, including loops.

Examples

# move through 4 points
for i range(5):
    movel(Pose(x=i*0.1))
# wait for a condition
while get_digital_in("start") == False:
    sync()

Conditions

Similarly, conditions are also supported.

Examples

if get_pathpilot_state() != "ready":
    notify("PathPilot is not ready", warning=True)

Subprograms

The TRPL supports subprograms / reusable Python functions.

Examples

def subprogram():
    movel(p[0, 100, 0, 0, 0, 0])

def main():
    subprogram()

Persistent Parameter Storage

The TRPL supports storing and retrieving Python objects to the persistent parameter store. This is useful if you want to store data between program runs.

robot_command.rpl.set_param(name: str, value: object) -> None

Sets a user parameter to a the defined value.

Parameters:

name – Parameter name.
value – Value to store.

set_param("my_waypoint", waypoint1)

robot_command.rpl.get_param(name: str, default: object = None) -> object

Fetch a stored user parameter.

Parameters:	name – Parameter name. default – Default value return if the parameter is not defined.
Returns:	Returns a base Python type, construct rpl types or returns a dict if the parameter is set, else returns the default value.

Examples

wp = get_param("my_waypoint", Pose())

robot_command.rpl.delete_param(name: str) -> None

Removes a user parameter

Parameters:

name – Parameter name to delete

delete_param("my_waypoint")

Probing

Under development

The probel command is a simple probing cycle:

robot_command.rpl.probel(target: Pose | Joints | str, a: float = 0.5, v: float = 0.1, v_retract: float = 0.1, away: bool = False, check_retract_contact: bool = False) -> Pose

Simple probing cycle that returns to the initial pose (regardless of the probe result). The sequence is:

Linear move at specified vel / accel scale towards the target position
Stop at probe contact, error condition, or motion end
Retract to original position
Raise any errors from the cycle, or return the probe result

Parameters:

target – end point of probing motion (probe cycle uses movel internally)
v – move velocity scaling factor 0.0 - 1.0
a – move acceleration scaling factor 0.0 - 1.0
v_retract – velocity scaling factor to use during retract phase
away – Probe towards work (default) if False, otherwise probe away from work
check_retract_contact – Optionally check for contacts during retract move (to avoid retracting into an obstacle and breaking a probe tip)

assumes mode 2/4 for probing, meaning an error will be thrown if it reaches the end without contact. Caller can catch this exception if they want mode 3/5 functionality

Examples

contact_pose = probel(probe_goal_pose, a=0.5, v=0.01, v_retract=0.1, away=False, check_retract_contact=False)

When called, probel() executes two motions:

movel() towards a target position at specified accelerations scale a (default 0.5), and velocity scale v (default 0.01), stopping when the probe makes contact.

movel() to retract to the original position (at velocity v_retract, default 0.1)

It’s also possible to create custom probing cycles by adding a “probe” keyword to any of the following move commands:

movej()
movel()
movec()

For example, a linear move that expects probe contact would look like this:

contact_type, contact_time, contact_joint_positions, contact_pose = movel(probe_goal_pose, probe=2)

The numerical value of the “probe” argument has the following meaning:

probe=2 means look for contact (rising edge), or raise a ProbeFailedError if it reaches the end without making contact, or if the probe is active at the start of the move.

probe=3 is like (2), but reaching the end of the motion is not an error.

probe=4 means look for the probe to break contact (falling edge). It raises a ProbeFailedError if it reaches the end without breaking contact, or if the probe is not active at the start.

probe=5 is like (4), but reaching the end of the motion is not an error.

probe=6 is for retraction. The motion will execute normally with the probe on or off, but if the probe signal gets a rising edge (i.e. retracting from a surface too far and hitting another surface), then it will raise a ProbeUnexpectedContactError

if the probe fails to make contact in mode 3 (or leave contact in mode 5), movel will return “None” instead of a result tuple.

Glossary of Probe Errors

There are several probing-specific errors that are reported as python exceptions. All of them are derived from the ProbeError exception class.

exception robot_command.rpl.ProbeUnexpectedContactError(error_code=None)

exception robot_command.rpl.ProbeContactAtStartError(error_code=None)

exception robot_command.rpl.ProbeFailedError(error_code=None)

Interrupts

The TRPL supports interrupts from external sources. This is useful for example to abort a program when a button connected to a digital input pin is pressed or to react to a message received from a ROS topic.

class robot_command.rpl.InterruptSource(value)

Interrupt source type.

DigitalInput: React to a change on a digital input pin.

UserIo: React to a change on a user IO pin. (triggered by HAL)

Program: React to a program interrupt. (e.g. a ROS topic)

robot_command.rpl.register_interrupt(source: InterruptSource, nr_or_name: int | str, fct: Callable) -> None

Registers a interrupt function to an interrupt source.

Parameters:

source – The interrupt source type.
nr_or_name – Number or name of the interrupt source, e.g. 1 for Digital Input 1.
fct – The function which should be called when the interrupt is triggered, if None is passed, this unregisters and disables the interrupt.

Examples

def interrupt_handler(value):
    if value:
        exit() # exit program when digital input 1 is high

register_interrupt(InterruptSource.DigitalInput, 1, interrupt_handler)

robot_command.rpl.trigger_interrupt(nr: int, value: Any) -> None

Triggers a program interrupt.

Parameters:

nr – Program interrupt number which should be triggered.
value – Value which is passed along with the triggered interrupt.

Examples

trigger_interrupt(2, 342.34)

Calibration

The robot_command.calibration module contains a few useful functions to write custom calibration programs. Note that these functions must be imported manually.

robot_command.calibration.calculate_tool_frame_4(wp1: Pose, wp2: Pose, wp3: Pose, wp4: Pose) -> Pose

Calculate the XYZ coordinates of the tool frame using 4 waypoints using the center of sphere method.

Parameters:	wp1 – Waypoint 1 wp2 – Waypoint 2 wp3 – Waypoint 3 wp4 – Waypoint 4
Returns:	The tool frame pose containing the XYZ tool frame.

robot_command.calibration.calculate_user_frame_3(origin_wp: Pose, x_axis_wp: Pose, y_axis_wp: Pose) -> Pose

Calculates the user frame origin pose based on 3 waypoints that lie on a plane.

The angular unit of all input poses must be in radians.

Parameters:	origin_wp – Origin of the plane. x_axis_wp – A waypoint that lies on the X-axis of the plane. y_axis_wp – A waypoint that lies in the direction of the Y-axis of the plane.
Returns:	Origin waypoint of the constructed user frame.

robot_command.calibration.calculate_user_frame_4(origin_wp: Pose, x_axis_wp: Pose, y_axis_wp: Pose, position_wp: Pose) -> Pose

Calculates the user frame origin pose based on 3 waypoints that lie on a plane and one additional waypoint which is used to define the origin location.

The angular unit of all input poses must be in radians.

Parameters:	origin_wp – Origin of the plane. x_axis_wp – A waypoint that lies on the X-axis of the plane. y_axis_wp – A waypoint that lies in the direction of the Y-axis of the plane. position_wp – Origin location of the new user frame.
Returns:	Origin waypoint of the constructed user frame.