Skip to main content

A high-level framework for programming robotic arms using TCP/IP connection.

Project description

Description

The arm-lightweight-controller is a very lightweight robotic arm controller for a real robot that is based on sending commands via tcp/ip (e.g. URScript commands in case of Universal Robots arms). It can be used as a standalone module - ROS is not required. Module is written in Python2.7 (ROS compatibility).

Difference between existing drivers (UR)

In ur_modern_driver the MoveIt is used to plan a path. In order to be usable in simulation, MoveIt planner computes many points between points specified in a trajectory. In arm-lightweight-controller there is no possibility to do that - path planning is performed entirely in the robot's controller. We cannot efficiently simulate the robot behavior, but it's fast and usable when it comes to the real robot.

TCP/IP communication details

Universal Robots provide the description of TCP/IP packages that are sent from robot's controller via sockets. The current communication schema in arm-lightweight-controller is based on that. In order to add more functionalities from this interface - see attached file under provided link.

Installation

A module arm-lightweight-controller can be installed using PyPi:

pip2 install robot_controller

Here is how to start. See examples/run.py file for an example of use.

import robot_controller
robot = robot_controller.Ur3("150.254.47.150", 30003, 30002)
...
# write your code here
...

In order to configure your Python environment just run the setup.py script.

Functionality

Short description of available methods in the Manipulator class. For more details please refer to the method descriptions in the code.

  • def move(self, trajectory, is_movej=True, is_pose=True, a=1, v=1, use_mapping=False) - send a move command to the specified robot. User can specify if it is a linear move / points are specified in a joint space or as poses / if the trajectory is in a robot's coordinate system or some external (see: set_mapping() method).
  • def get_pose(self) - reads current pose from a byte stream.
  • def get_joints(self) - reads current joint coordinates from a byte stream.
  • def set_mapping(self, matrix) - sets a mapping between a robot coordinate system and some external one. While this method is invoked, user can pass trajectory in some external coordinate system (see: move() command).
  • def execute_in_force_mode(self, trajectory_commands, task_frame, selection_vector, wrench, type=1, limits) - Executes constructed commands in the specified force mode. Please see URScript API reference doc - force_mode()

Contribution

All questions to: michal.bednarek@put.poznan.pl

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

robot_controller-0.3.tar.gz (10.0 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page