Master's Degree Thesis Proposal

Student: Ing. Israel Chaves Arbaiza, B99336


ARCOS-Lab's Humanoid Robot

Scenarios

Omnidirectional Mobile Platform

  • Mecanum wheels
  • Contains the electrical power control for the robot
  • Rotational joint at the back
  • Our own designed PCB's, OpenCoRoCo, to power control

Torso

  • Uses a power screw to move up and down the "shoulders" of the robot
  • The "shoulders" were designed as a result of simulation
  • 2 KUKA LWR4+ arms and 2 DLR HIT II hands
  • Contains the computers for each subsystem of the robot

Control suave

<iframe width="800" height="400" src="media/impedance_control.mp4" title="Video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>

Head and emotional face

  • The neck moves using a pan-tilt system
  • Head with two RealSense sensors to detect the 3D space
  • Two profesional lenses Sony
  • One thermal camera

Literature Review

Capturing robot workspace structure: representing robot capabilities

  • Intelligent Robots and Systems 2007
  • Authors: Zacharias F., Borst, C. and Hirzinger, G.
  • Capability maps concept creation, it contains directional structure

Smart placement of a two-arm assembly for an everyday object manipulation humanoid robot based on capability maps

  • IWOBI 2018
  • Authors: Chaves I., García D. and Ruiz, F.
  • Capability maps to calculate a score for different two-arm assembly placements
  • It's center in getting the best dual arm setup for collaboration

Efficient Coverage of 3D Environments with Humanoid Robots Using Inverse Reachability Maps

  • Humanoids 2017
  • Authors: Oßwald, S.; Karkowski, P. and Bennewitz, M.*
  • They use Reachability Maps instead of Capability Maps
  • Center in the vision of the environment
  • Intelligent Robots and Systems 2009
  • Authors: Stulp, F.; Fedrizzi, A. and Beetz, M.
  • They use Probabilistic Distributions
  • The goal is to get the best base position

Combining Navigation and Manipulation Costs for Time-Efficient Robot Placement in Mobile Manipulation Tasks

  • IEEE Robotics and Automation Letters 2022
  • Authors: Reister, F.; Grotz, M.; and Asfour, T.
  • They use reachability maps for both arms
  • Tasks using just one arm

Objectives

General Objective:

Develop a method to create multiple cognitive capability maps, for a specific robot in a particular scenario

Specific Objectives:

  • Create a Cognitive Capability Map (CCM) for a complete scenario, that includes objects and elements in the room
  • Implement the ability to generate multiple capability maps for every combination of one hand, two hands and the robot's perception system
  • Generate a list of manipulation tasks with every combination of one hand, two hands and the vision system, for both a kitchen and a retails scenario.
  • Develop a library with the cognitive capability maps of every combination for the robot in the scenario
  • Implement a system that chooses that using the library and a motion planner, chooses the best CCM suitable for every task, in the particular scenario.
  • Elaborate experiments that verify the efficiency of the cognitive capability mapping method for both scenarios and theirs chosen tasks

Questions?




<img width="900" height="260" src="media/scenarios.png">