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
http://docs.ros.org/kinetic/api/moveit_tutorials/html/doc/move_group_interface/move_group_interface_tutorial.html#cartesian-paths
I would recommend you to check this page.
Asked by saikishor on 2018-07-23 11:54:16 UTC
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