CommandTalk: A Spoken-Language Interface to a Battlefield Simulator

John Dowding, SRI International and Robert Moore, NASA Ames Research Center
moore@riacs.edu, dowding@ai.sri.com

Seminar on People, Computers, and Design
Stanford University October 9, 1998

 

CommandTalk is a spoken-language interface to synthetic forces in entity-based battlefield simulations, developed by SRI International over the last 3 years. The principal goal of CommandTalk is to let commanders interact with simulated forces by voice in a manner as similar as possible to the way they would command actual forces. CommandTalk was initially deployed in LeatherNet, a simulation and training system for the Marine Corps developed under direction of the of the Naval Command, Control and Ocean Surveillance Center, RDT&E Division (NRaD). CommandTalk was later adapted to support all 4 services, and used in support of a large distributed joint exercise as part of DARPA's STOW program. 

This talk will focus on the technical aspects of the CommandTalk architecture, on design decisions made while applying spoken-language technology to a large legacy application, and on some of the user's responses to those decisions. While no formal evaluation of CommandTalk was carried out, we gathered extensive anecdotal evidence regarding user preferences and user acceptance.

John Dowding is a Senior Computer Scientist in the Natural Language group of the Artificial Intelligence Center at SRI International. John has been the lead engineer designing the Gemini spoken-language understanding system and its application in CommandTalk, a spoken-language interface to a battlefield simulator, and principal investigator on several related projects. Before joining SRI in 1990, John was a researcher at Unisys Corporation in the Logic-Based System group, working on parsing and related issues for the Pundit text-processing project.

 Robert C. Moore is the Director of the Research Institute for Advanced Computer Science (RIACS), located at NASA Ames Research Center in Mountain View, California. Before joining RIACS in 1998, Dr. Moore spent 20 years in artificial intelligence research at SRI International, where at various times he served as a Principal Scientist in SRI's Artificial Intelligence Center, Director of SRI's Natural-Language Research Program, and founding Director of SRI's Computer Science Research Centre in Cambridge, England. Dr. Moore has also been a Fellow of the Center for Advanced Study in the Behavioral Sciences, a Visiting Industrial Lecturer at Stanford University, and a Fellow Commoner of Churchill College, Cambridge. Areas of Dr. Moore's research have included natural-language processing, speech understanding, knowledge representation, and automated reasoning. He has been a Fellow of the American Association for Artificial Intelligence since 1991, and has served on the editorial boards of the journals "Artificial Intelligence" and "Computational Linguistics". His numerous publications include the recent book "Logic and Representation," published by Stanford University's Center for the Study of Language and Information in 1995.

 

Titles and abstracts for all years are available by year and by speaker.

For more information about HCI at Stanford see