Robotics StackExchange | Archived questions

Correct implementation of computeCartesianPath?

Hello everyone. I'm quite new to ROS, doing a practicum project right now and have been pretty stumped with this for a few days. I have a TIAGO Robot, with a 6DOF arm with whom I want to be able to move on a straight line. The "straight" part is important as we'll be having a pressure sensor on the fingertip with which we want to do "feeling" of surfaces, so we need to be able to move it on a straight line (maintaining orientation of it) on the XY plane.

So I've been basically doing some frankencode because while I'm familiar with C++, I'm not with ROS and MoveIt!; and been piecing something together from a basic code in the TIAGO tutorials that moves the actuator to a cartesian point and orientation (but not necessarily in a straight line) and the moveit tutorial that deals with Cartesian path. Current code:

   /*
 * Software License Agreement (Modified BSD License)
 *
 *  Copyright (c) 2016, PAL Robotics, S.L.
 *  All rights reserved.
 *
 *  Redistribution and use in source and binary forms, with or without
 *  modification, are permitted provided that the following conditions
 *  are met:
 *
 *   * Redistributions of source code must retain the above copyright
 *     notice, this list of conditions and the following disclaimer.
 *   * Redistributions in binary form must reproduce the above
 *     copyright notice, this list of conditions and the following
 *     disclaimer in the documentation and/or other materials provided
 *     with the distribution.
 *   * Neither the name of PAL Robotics, S.L. nor the names of its
 *     contributors may be used to endorse or promote products derived
 *     from this software without specific prior written permission.
 *
 *  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
 *  "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
 *  LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
 *  FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
 *  COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
 *  INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
 *  BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
 *  LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
 *  CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 *  LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
 *  ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 *  POSSIBILITY OF SUCH DAMAGE.
 */

/** \author Jordi Pages. */

// ROS headers
#include <ros/ros.h>
#include "shape_msgs/SolidPrimitive.h"
#include "moveit_msgs/BoundingVolume.h"


// MoveIt! headers
#include <moveit/move_group_interface/move_group_interface.h>
//#include <moveit_visual_tools/moveit_visual_tools.h>

// Std C++ headers
#include <string>
#include <vector>
#include <map>
#include <cmath>

int main(int argc, char** argv)
{
  ros::init(argc, argv, "plan_arm_torso_ik_idril");

  if ( argc < 7 )
  {
    ROS_INFO(" ");
    ROS_INFO("\tUsage:");
    ROS_INFO(" ");
    ROS_INFO("\trosrun tiago_moveit_tutorial plan_arm_torso_ik_idril  x y z  r p y");
    ROS_INFO(" ");
    ROS_INFO("\twhere the list of arguments specify the target pose of /arm_tool_link expressed in /base_footprint");
    ROS_INFO(" ");
    return EXIT_FAILURE;
  }

  geometry_msgs::PoseStamped goal_pose;
  goal_pose.header.frame_id = "base_footprint";
  goal_pose.pose.position.x = atof(argv[1]);
  goal_pose.pose.position.y = atof(argv[2]);
  goal_pose.pose.position.z = atof(argv[3]);
  goal_pose.pose.orientation = tf::createQuaternionMsgFromRollPitchYaw(atof(argv[4]), atof(argv[5]), atof(argv[6]));

  ros::NodeHandle nh;
  ros::AsyncSpinner spinner(1);
  spinner.start();

  std::vector<std::string> torso_arm_joint_names;
  //select group of joints
  moveit::planning_interface::MoveGroupInterface group_arm_torso("arm_torso");
  const robot_state::JointModelGroup* joint_model_group =
    group_arm_torso.getCurrentState()->getJointModelGroup("arm_torso");

  //choose your preferred planner

  group_arm_torso.setPlannerId("SBLkConfigDefault");
  group_arm_torso.setPoseReferenceFrame("base_footprint");
  group_arm_torso.setPoseTarget(goal_pose);

  ROS_INFO_STREAM("Planning to move" <<
                  group_arm_torso.getEndEffectorLink() << " to a target pose expressed in " <<
                  group_arm_torso.getPlanningFrame());

  group_arm_torso.setStartStateToCurrentState();
  group_arm_torso.setMaxVelocityScalingFactor(1.0);

  // Getting Basic Information
  // ^^^^^^^^^^^^^^^^^^^^^^^^^
  //
  // We can print the name of the reference frame for this robot.
  ROS_INFO_NAMED("tutorial", "Reference frame: %s", group_arm_torso.getPlanningFrame().c_str());

  // We can also print the name of the end-effector link for this group.
  ROS_INFO_NAMED("tutorial", "End effector link: %s", group_arm_torso.getEndEffectorLink().c_str());

  // Planning with Path Constraints
  // ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  //
  // Path constraints can easily be specified for a link on the robot. straight line
  // Let's specify a path constraint and a pose goal for our group.
  // First define the path constraint.
  ROS_INFO_STREAM("Calculating Restrictions");
  geometry_msgs::PoseStamped start_pose;
  start_pose.header.frame_id = "base_footprint";
  start_pose.pose = group_arm_torso.getCurrentPose().pose;

  geometry_msgs::PoseStamped vector_direccional;
  vector_direccional.pose.position.x = goal_pose.pose.position.x - start_pose.pose.position.x;
  vector_direccional.pose.position.y = goal_pose.pose.position.y - start_pose.pose.position.y;
  vector_direccional.pose.position.z = goal_pose.pose.position.z - start_pose.pose.position.z;

  double dist = sqrt(pow(vector_direccional.pose.position.x,2) + pow(vector_direccional.pose.position.y, 2) + pow(vector_direccional.pose.position.z,2));

  ROS_INFO_STREAM("Calculated directional vector: " <<
                  vector_direccional.pose.position <<
                  "Start and goal points are at distance: " << dist);

  moveit_msgs::OrientationConstraint ocm;
  ocm.link_name = "arm_tool_link";
  ocm.header.frame_id = "base_footprint";
  ocm.orientation = group_arm_torso.getCurrentPose().pose.orientation;
  ocm.absolute_x_axis_tolerance = 0.1;
  ocm.absolute_y_axis_tolerance = 0.1;
  ocm.absolute_z_axis_tolerance = 0.1;
  ocm.weight = 1.0;
  ROS_INFO_STREAM("Calculated Orientation Constraint");

  moveit_msgs::Constraints test_constraints;
  test_constraints.orientation_constraints.push_back(ocm);
  group_arm_torso.setPathConstraints(test_constraints);
  ROS_INFO_STREAM("Constraints set");

  //Cartesian Path
  std::vector<geometry_msgs::Pose> waypoints;
    waypoints.push_back(start_pose);
    waypoints.push_back(goal_pose);
  }

  moveit_msgs::RobotTrajectory trajectory;
  const double jump_threshold = 100;
  const double eef_step = dist;
  double fraction = 0;

  double old_jump_threshold = jump_threshold;
  double old_eef_step = eef_step;
  double good_jump_threshold = jump_threshold;
  double good_eef_step = eef_step;
  double good_fraction = fraction;

  //This loop is only to check how eef_step and jump_treshold affect the outcome.
  for(int i = 1; i <= 300; ++i){
      for(int j = 1; j <= 1000; ++j){
               fraction = group_arm_torso.computeCartesianPath(waypoints,
                                                               eef_step/i,
                                                               jump_threshold/j,
                                                               trajectory,
                                                               true);
      if (fraction < 0.15) break;
      if (fraction > good_fraction){
        ROS_INFO_STREAM("Found better parameters: " << good_fraction* 100.0 << '%' << " ==> " << fraction * 100.0 << '%' << ":");
        ROS_INFO_STREAM("eef " << old_eef_step << " ==> " << good_eef_step);
        ROS_INFO_STREAM("jump " << old_jump_threshold << " ==> " << good_jump_threshold);
        good_fraction = fraction;
        old_eef_step = good_eef_step;
        old_jump_threshold = good_jump_threshold;
        good_eef_step = eef_step/i;
        good_jump_threshold = jump_threshold/j;

      }
      if (fraction == 1) break;
    }

    if (fraction == 1) break;
  }
  if (fraction == 1) ROS_INFO_NAMED("tutorial", "Visualizing plan 4 (Cartesian path) (%.2f%% achieved)", fraction * 100.0);
  else{
    fraction = group_arm_torso.computeCartesianPath(waypoints,
                                                    good_eef_step,
                                                    good_jump_threshold,
                                                    trajectory,
                                                    true);
  ROS_INFO_NAMED("tutorial", "Best I could do was (%.2f%% achieved)", fraction * 100.0);
  ROS_INFO_STREAM("With eef = " << good_eef_step << ", jump = " << good_jump_threshold);

  }

  moveit::planning_interface::MoveGroupInterface::Plan my_plan;
  //set maximum time to find a plan
  ROS_INFO_STREAM("Finding Plan...");
  group_arm_torso.setPlanningTime(20.0);
  bool success = bool(group_arm_torso.plan(my_plan));

  if ( !success )
    throw std::runtime_error("No plan found");

  ROS_INFO_STREAM("Plan found in " << my_plan.planning_time_ << " seconds");

  // Execute the plan
  ros::Time start = ros::Time::now();

  moveit::planning_interface::MoveItErrorCode e = group_arm_torso.move();
  if (!bool(e))
    throw std::runtime_error("Error executing plan");

  ROS_INFO_STREAM("Motion duration: " << (ros::Time::now() - start).toSec());
  group_arm_torso.clearPathConstraints();
  spinner.stop();

  return EXIT_SUCCESS;
}

I realise that it's probably just a newbie error somewhere but I've been scouring the forums about the ComputeCartesianPath() function and I don't really get how it works. Sometimes it returns 100% but the path comes out clearly curved, other times it's 0% but the path is visually straight. To add to this, is there a way to "compare" the outcome of the planned (executed) path to the "perfect" path? To make sure it's actually straight and have a number that gives me the error.

Thank you very much for your help

Asked by idrilirdi on 2018-07-23 09:08:46 UTC

Comments

I'm familiar with that page, it's where I started. It still doesn't solve my problem, sorry. Sometimes it returns 100% but the path comes out clearly curved, other times it's 0% but the path is visually straight. I can't find a correlation between the fraction returned and the results I get

Asked by idrilirdi on 2018-07-24 02:45:45 UTC

Answers