ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question

P Tiago P's profile - activity

2017-11-29 05:42:00 -0500 received badge  Taxonomist
2016-07-09 21:41:34 -0500 received badge  Famous Question (source)
2016-01-25 23:46:06 -0500 received badge  Famous Question (source)
2015-08-07 07:41:34 -0500 commented question obstacles are not removed from the costmap

@David: I updated the screenshots

2015-08-06 17:07:13 -0500 received badge  Famous Question (source)
2015-08-06 11:17:07 -0500 received badge  Editor (source)
2015-07-30 06:34:28 -0500 received badge  Notable Question (source)
2015-07-27 12:23:24 -0500 received badge  Popular Question (source)
2015-07-26 11:24:44 -0500 commented question obstacles are not removed from the costmap

@David: Thanks for asking, I've updated the post.

2015-07-25 17:49:01 -0500 asked a question obstacles are not removed from the costmap

Hi,

I've searched and i found similar questions, but no conclusions.

scenario:

ROS: Indigo

Laser: XV11

The robot is stable not moving, i move near the robot, and the map is updated with an obstacle, i move away, and the obstacle is not removed (unless the move_base is restarted), the laser is clear, the obstacle is still there plus some yellow dots.

rviz snapshots:

initial: https://dl.dropboxusercontent.com/u/1...

obstacle present: https://dl.dropboxusercontent.com/u/1...

obstacle gone: https://dl.dropboxusercontent.com/u/1...

obstacle gone (laser only): https://dl.dropboxusercontent.com/u/1...

If i go again to the front of the robot the obstacle area get's bigger, and stays there even when i move away.

Any clues ?

Regards,

Tiago

Update1: (costmap parameters)

==========================================================

robot_radius: 0.22545

inflation_layer:
  enabled: true
  cost_scaling_factor: 10.0
  inflation_radius: 0.25

obstacle_layer:
  enabled: true
  obstacle_range: 2.5
  raytrace_range: 3.0
  max_obstacle_height: 0.6
  min_obstacle_height: 0.0
  observation_sources: scan_source
  scan_source:
    data_type: LaserScan
    topic: /xv11
    marking: true
    clearing: true
    expected_update_rate: 0.4
    observation_persistence: 0.0

==========================================================

local_costmap:
   plugins:
      - {name: sonar_source, type: "range_sensor_layer::RangeSensorLayer"}
      - {name: obstacle_layer, type: 'costmap_2d::ObstacleLayer'}
      - {name: inflation_layer, type: 'costmap_2d::InflationLayer'}
   global_frame: /odom
   robot_base_frame: /base_footprint
   update_frequency: 3.0
   publish_frequency: 1.0
   static_map: false
   rolling_window: true
   transform_tolerance: 1.0
   map_type: costmap
   width: 6.0
   height: 6.0
   resolution: 0.01

==========================================================

global_costmap:
  plugins:
    - {name: sonar_source, type: "range_sensor_layer::RangeSensorLayer"}
    - {name: static_layer, type: 'costmap_2d::StaticLayer'}
    - {name: obstacle_layer, type: 'costmap_2d::ObstacleLayer'}
    - {name: inflation_layer, type: 'costmap_2d::InflationLayer'}
  global_frame: /map
  robot_base_frame: /base_footprint
  update_frequency: 1.0
  publish_frequency: 0.0
  static_map: true
  rolling_window: false
  resolution: 0.01
  transform_tolerance: 1.0
  map_type: costmap
2015-07-25 17:13:36 -0500 received badge  Enthusiast
2015-07-14 21:20:14 -0500 received badge  Scholar (source)
2015-07-13 07:18:20 -0500 received badge  Notable Question (source)
2015-07-12 10:41:23 -0500 received badge  Popular Question (source)
2015-07-11 18:21:43 -0500 asked a question teleop + Collision detection

What is the most effective way to implement teleop (twist) with collision detection, basically how to query the costmap to determine if is clear of free to send the twist msg ?

Cheers, Tiago

2015-06-11 04:04:20 -0500 received badge  Notable Question (source)
2015-03-10 13:45:24 -0500 received badge  Popular Question (source)
2015-03-09 14:51:04 -0500 asked a question razor_imu_9dof stick orientation

The stick version the G/M/A axes are not aligned and are inverted (bottom board)

https://cdn.sparkfun.com/assets/parts...

versus the board:

https://cdn.sparkfun.com/assets/parts...

Firmware axes adjustments (sensors.ino)

For all the boards: A=> x = y; y=x; z=z; G=> x = -y; y =-x; z = -z;

For the 10724: M=> x = x; y = -y; z=-z;

1) What should be the stick alignment ?

if both boards are aligned horizontal text:

Board: x (down) / y (right) / z (counterclockwise)

stick: invert horizontal axes(bottom print) we have:

G=> x (up) / y(left) / z (counterclockwise) A=> y (left) / x(down) M=> x(left) / y(down)

and looking to the firmware the only existent adjustment (HW__VERSION_CODE == 10724) is done over the magnet readings, but the Gyro and Accel are not aligned, shouldn't the firmware adjust those two ?

Regards,

Tiago