AdaptiveCollisionHand-Code

Code


Overview

The experiment makes use of the OpenRAVE simulation environment, the YARP communications protocol and the open 'spacenavd' driver (from the Spacenav project) for the 3DConnexion SpaceNavigator input device.

The total effort required to install and prepare the experiment on a "clean" Ubuntu machine is about 1-2 hours.


Tested platforms

  • Linux Ubuntu 10.04, 32 bit
  • Linux Ubuntu 11.04, 64 bit

Instructions


Architecture


Modules

The modules used are detailed below:

  • Spacenavigator
    • YARP driver that connects to the 'spacenavd' driver for the SpaceNavigator input device.
    • Executable: 'spacenavigator_spnav'
  • Experiment manager
    • The experiment manager handles the experiment logic, the trial order, and stores the trial-level and trajectory results.
    • Executable: 'main_exp_manager_steering'
  • Trajectory player
    • This module is used to play a trajectory file and output it through a port. Here used for playing noise for simulating a generic disability. See the experimental paper description.
    • Executable: 'main_trajectory_player'
  • Window feedback
    • Provides feedback on the time for the current attempt and the state of the experiment (through color changes) to the participant.
    • Executable: 'main_window_feedback'
  • Adaptive shared control
    • Performs the adaptive shared control. See the experimental paper description.
    • Executable: 'main_uci_proximity_neural'
  • Neural network weight storage
    • Receives, sends and stores to file the neural network weights for a given participant.
    • Executable: 'kdb_neural_weights'
  • OpenRAVE driver
    • Provides the virtual environment for the experiment, based on a real kitchen. See the experimental paper description.
    • Executable: 'exp_openrave_driver'

[Home] [ExperimentalPapers] [AdaptiveCollisionHand]


Related

Wiki: AdaptiveCollisionHand-Code-Install
Wiki: AdaptiveCollisionHand-Code-Startup
Wiki: AdaptiveCollisionHand
Wiki: ExperimentalPapers
Wiki: Home

MongoDB Logo MongoDB