Bin-picking

My Master's Thesis Blog

Precise Bin-picking using a 3D Sensor and a 1D Laser Sensor

Bin-picking is the name of the technique used by a robot to grasp objects that are randomly disposed inside a box or on a pallet.

My purpose with this thesis is to develop a solution for a different process of bin-picking using a 3D sensor to generate a point cloud of the objects, and a 1D laser sensor to correctly determinate the exact coordinates of the grasping point.

Week 0 | (19/02 - 25/02)

  • ROS Installation and tutorials.


Week 1-2 | (26/02 - 11/03)

  • ROS Workshop.
  • Preliminary Report.


Week 3 | (12/03 - 18/03)

  • Presentation for the first LAR Meeting.
  • Research for the state of the art.
  • First data acquisition from the Kinect.


Week 4 | (19/03 - 25/03)

  • Continuation of the research for the state of the art.
  • Developments on the installation of the laser sensor.
  • Conduction of a few tests for the different installation positions of the Kinect in the robot.
  • Segmentation of the data acquired from the Kinect.


Week 5-6 | (26/03 - 5/04)

  • Installation of the laser sensor and processing its output signal.
  • Calibration of the Kinect.
  • Processing of the data acquired from the Kinect to determine the centroid of an object and its normal.


Week 7 | (6/04 - 12/04)

  • Support for the Kinect.
  • Changes on the circuit to filter the laser sensor readings.
  • Beginning of the calibration of all the hardware utilized.


Week 8 | (13/04 - 19/04)

  • Calibration of the Kinect and the laser sensor.


Week 9 | (20/04 - 26/04)

  • Point Cloud Segmentation.


Week 10 | (26/04 - 2/05)

  • Evaluation of the Kinect calibration.


Week 11 | (2/05 - 10/05)

  • Correct calibration of the Kinect.


Week 12-13-14-15 | (21/05 - 13/06)

  • System Integration.


Week 16 | (21/06 - 28/06)

  • Tests.


Week 17 | (2/07 - 9/07)

  • Demonstration.

Demonstration of the developed Bin-picking process.