You are here

Voice Activated Cockpit Management Systems: Voice-Flight NexGen

Award Information
Agency: National Aeronautics and Space Administration
Branch: N/A
Contract: NNX13CL13P
Agency Tracking Number: 125604
Amount: $125,000.00
Phase: Phase I
Program: SBIR
Solicitation Topic Code: A1.03
Solicitation Number: N/A
Timeline
Solicitation Year: 2012
Award Year: 2013
Award Start Date (Proposal Award Date): 2013-05-23
Award End Date (Contract End Date): 2013-11-23
Small Business Information
CA
Carlsbad, CA 92011-4404
United States
DUNS: 968727755
HUBZone Owned: No
Woman Owned: Yes
Socially and Economically Disadvantaged: No
Principal Investigator
 Doinita Serban
 Principal Investigator
 (619) 917-4299
 serban@southernairaviation.com
Business Contact
 Doinita Serban
Title: Business Official
Phone: (619) 917-4299
Email: serban@southernairaviation.com
Research Institution
 Stub
Abstract

Speaking to the cockpit as a method of system management in flight can become an effective interaction method, since voice communication is very efficient. Automated Speech Recognition in general, and aided and abetted now by Southern Air Aviation's Voice Activated Cockpit Management Systems including the innovative heuristic technology Voice-Flight NexGen (VFNG) has advanced significantly in recent years and is now ideally suited to this application. VFNG is speaker-independent, achieves a word recognition rate of 98% in high noise and up to 120Db, and permits the chaining of up to four commands in a single utterance and a correction of a misspoken command in the same utterance. All of these performance requirements are met within a 2% time-slice of a 1.2GHz Power PC processor and within 1 GB of memory. VFNG includes the following unique key capabilities, Grammar Development, Dictionary Development, Automated Batch Grammar Testing, Grammar Coverage Query, Phonetic Distance Analysis, and Co?articulation Handling, Command Based Confidence Algorithm, Accent Tolerability Handling, State Based Dynamic Grammar, and Clipped Audio Recovery Processing. VFNG incorporates proprietary Julius based grammar and vocabulary of advanced performance and reduced recognition error together with an acoustic model derived from the VoxForge acoustic model. The VFNG-Emergency Failures (VFNG-E) system is used as the sole human/machine interface in flight in real emergency situations, in addition to the system having applications for simulator and ground training. This training system is remarkable in that it achieves very high recognition rates (98%) with a very large command set (131,000 unique words and thousands of word combinations) in challenging environmental conditions and operator profiles providing much added safety and efficiency to the pilot in the cockpit.

* Information listed above is at the time of submission. *

US Flag An Official Website of the United States Government