Control of large numbers of robots

Proposed paper snippets

  • In contrast to other approaches, I propose to examine the game interface in terms of interactions with the other agents present.
  • Scales to operator skill - provides intuitive gui/mouse based and provides hotkeys to let experienced users do things more quickly.
  • We restrict our view to i/o devices found on the common computer system. We neglect issues such as cameras, and input output devices. For a treatment of the topic see [this] excellent paper.
  • Focused on agent interaction, not interface. other papers deal with interface.
  • Previous papers have explored applicability of RTS interface to robot interaction and the design of the user interface itself. My focus is on interaction with the agents inside the game.

Video game-based framework for HRI

  • Broad survey of video game techniques
  • Applied specifically to UAVs, but goal is general framework
  • A human interaction focus - segments features into “inputs” and “outputs” - different from my proposed approach
  • Good bibliography for my research area… spend time reading here.

Multi-robot control based on RTS interaction paradigm

  • Paradigm works, but must be adapted
  • Currently two categories of interfaces: single complex robot and distributed robot system. Need to be able to control multiple complex robots
  • “complex” - ability to perform extended missions that call for more than one kind of task.
  • Many current systems operate in “hands off” or full tele-operated control.
  • Footnote 25, 27 - human robot teams. 17 - Boeing study
  • “user centered design”
  • observed SWAT teams - good model for human-human interaction in stressful situation, now integrate in robots
  • Complete world knowledge not realistic in real world environment
  • NOTE: maybe I should also look at rpg-ish interfaces… that tend to facilitate “put rock in box” types of commands.
  • NOTE: one of the fundamental challenges in multi-robot control is to put all robots in a coherent space (same coordinate system) where the user can visually see relations between them and interact with them.
  • Good results
  • Results indicated that the user Trusted the robot more with a better interface.

Speaking Swarmish

  • Short paper - 4 pages
  • MIT 2006
  • Large (116) swarm of robots.
  • Deals with develop-run-debug cycle on a local swarm. Emphasizes ways for the operator to easily determine the robots state without consulting a computer screen.
  • Applications to human-robot interaction. How can a human easily see the internal state of a robot?
  • Applies strategies from RTS games. Mentions that use of this style of interface to control robots has mixed results (easier to move them or type commands). Mentions that such interfaces are very useful for information display.
  • Primarily deals with “Swarmish”, a “communication protocol” that the robots use to communication information back to humans.
    • Blinking LEDs. Seem to work best when using all LEDs or one at a time. Multiple independent signals are hard to read.
      • Uses two intensities and both square and sine waves
      • Combination gives 108 patterns, all readable in less than 1/2 second.
    • Audio signals
      • inspired by old school debugging using radios
      • makes combinations of states intelligible. Can quickly spot “discordance” and id bugs.
      • Users especially sensitive to demo and rhythm.
  • looking at augmented reality for future.
hcmi/brian_pendleton_reading_notes.txt · Last modified: 2014/08/13 16:16 by tlund1
Back to top
CC Attribution-Share Alike 4.0 International = chi`s home Valid CSS Driven by DokuWiki do yourself a favour and use a real browser - get firefox!! Recent changes RSS feed Valid XHTML 1.0