# US 12,118,982 B2 · System and method for constraining air traffic communication (ATC) transcription in real-time > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,118,982 B2 - Document ID: US12118982B2 - Status: Granted - Kind: B2 - Application number: 17/664,756 - Filed: 2022-05-24 - Published: 2023-10-12 - Granted: 2024-10-15 - Earliest claimed: 2022-04-11 - Assignees: Honeywell International Inc. - Applicants: Honeywell International Inc. - Inventors: Jitender Agarwal; Mohan M Thippeswamy - Web page: https://patentgrep.com/patent/US12118982B2 ## CPC - G10L 15/02: Feature extraction for speech recognition; Selection of recognition unit - G06F 40/279: Recognition of textual entities - G08G 5/21: located onboard the aircraft - G08G 5/26: Transmission of traffic-related information between aircraft and ground stations - G08G 5/34: for flight plan modification - G08G 5/53: for cruising - G10L 15/193: Formal grammars, e.g. finite state automata, context free grammars or word networks - G10L 15/26: Speech to text systems (G10L15/08 takes precedence) - G10L 15/32: Multiple recognisers used in sequence or in parallel; Score combination systems therefor, e.g. voting systems - G06F 40/242: Dictionaries - G08G 5/00: Traffic control systems for aircraft - G08G 5/55: for a single aircraft - G10L 2015/228: {of application context} - G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G08G: TRAFFIC CONTROL SYSTEMS (guiding railway traffic, ensuring the safety of railway traffic B61L; arrangement of road signs or traffic signals E01F9/00; radar or analogous systems, sonar systems, lidar systems specially adapted for traffic control G01S13/91, G01S15/88, G01S17/88; {radar or analogous systems, sonar systems, lidar systems specially adapted for anti-collision purposes G01S13/93, G01S15/93, G01S17/93}) ## Abstract Systems and methods are provided for the selection of a speech model for automatic speech recognition during the runtime of a transcription system, the system includes an event detector to determine one of a number of flight events that include flight plan changes and phase transitions based on data received from a set of inputs; an intelligent keyword generator to collate a set of keywords associated with the flight plan information and to generate a wordlist in response to a determination by the event detector of flight plan changes or flight phase transitions; and a processor to determine whether the wordlist is covered by a current speech model implemented in the automatic speech recognition wherein if the wordlist is not covered by the current speech model, then the processor to select a pre-built speech model that covers the wordlist for use as the current speech model in the automatic speech recognition. ## Claims ### Claim 1 A transcription system with a selectable speech model used in automatic speech recognition comprising: an event detector configured to determine one of a number of flight events that comprise flight plan changes and a flight phase transition based on event data received from a set of inputs; an intelligent keyword generator in operable communication with the event detector and configured to collate a set of keywords associated with at least flight plan information to generate a wordlist in response to a determination by the event detector of the flight plan changes or the flight phase transition wherein the wordlist contains keywords associated with the flight plan changes and the flight phase transition; and a processor in operable communication with the intelligent keyword generator and configured to determine, based on the wordlist from the intelligent keyword generator, whether the wordlist is covered by a current speech model implemented in the automatic speech recognition of the transcription system, wherein if the wordlist is not covered by the current speech model, then the processor is further configured to communicate with a database storing one or more pre-built speech models to select a pre-built speech model that covers the wordlist for use as the current speech model in the automatic speech recognition of the transcription system. ### Claim 2 (depends on 1) The transcription system of claim 1, further comprising: if the pre-built speech model is not available that covers the wordlist from the intelligent keyword generator, then the processor is further configured to generate a new speech model for use as the current speech model with the automatic speech recognition of the transcription system to enable coverage of the wordlist by the current speech model and constraint of the current speech model to at least the flight plan changes or the flight phase transition. ### Claim 3 (depends on 2) The transcription system of claim 2, wherein the processor is further configured to: determine the coverage of the wordlist from the intelligent keyword generator by comparison of the keywords in the wordlist contained in each of the one or more pre-built speech models stored in the database. ### Claim 4 (depends on 3) The transcription system of claim 3, wherein the intelligent keyword generator is further configured to: collate the keywords used in communication in the flight plan changes or the flight phase transition for comparison of keyword coverage in each of the one or more pre-built speech models stored in the database. ### Claim 5 (depends on 4) The transcription system of claim 4, wherein the event detector is further configured to: receive the event data from the set of inputs, wherein the set of inputs comprises a first input of pilot input, a second input of data input from one or more aircraft systems that include Flight Management System (FMS) data, a third input of Air Traffic Control (ATC) clearance data, and a fourth input of log data generated by one or more checklists or pilot logs. ### Claim 6 (depends on 5) The system of claim 5, wherein the set of inputs are configured in a hierarchy by the processor to determine a flight phase change or flight phase transition with the pilot input given a highest value. ### Claim 7 (depends on 6) The transcription system of claim 6, wherein the processor is further configured to: implement a plurality of checks to determine whether data from the first, second, third, or fourth input triggers the determination by the event detector of the flight plan change or the flight transition for re-selecting of the current speech model in use by the transcription system. ### Claim 8 A method of implementing automatic speech recognition during runtime of a transcription system, the method comprising: determining, by an event detector, one of a number of flight events that comprise flight plan changes and a flight phase transition based on event data received from a set of inputs; collating, by an intelligent keyword generator, a set of keywords associated with at least flight plan information for generating a wordlist in accordance with the flight plan changes or the flight phase transition determined by the event detector, wherein the wordlist contains keywords associated with the flight plan changes and the flight phase transition; determining, by a processor based on the wordlist from the intelligent keyword generator, whether the wordlist is covered by a current speech model implemented in automatic speech recognition of the transcription system; and in response to a determination that the wordlist is not covered by the current speech model, selecting, by the processor by communicating with a database storing one or more pre-built speech models, the pre-built speech model that covers the wordlist. ### Claim 9 (depends on 8) The method of claim 8, further comprising: in response to the determination that the pre-built speech model covering the wordlist from the intelligent keyword generator is not available, generating by the processor, a new speech model for use as the current speech model for the automatic speech recognition of the transcription system for enabling coverage of the wordlist by the current speech model and for constraining the current speech model to at least the flight plan changes or the flight phase transition. ### Claim 10 (depends on 9) The method of claim 9, further comprising: determining coverage, by the processor, of the wordlist from the intelligent keyword generator by comparison of the wordlist to the keywords contained in each of the one or more pre-built speech models stored in the database. ### Claim 11 (depends on 10) The method of claim 10, further comprising: collating, by the processor, keywords of the wordlist used in communication in the flight plan changes or the flight phase transition for comparison of keyword coverage in each of the one or more pre-built speech models stored in the database. ### Claim 12 (depends on 11) The method of claim 11, further comprising: receiving, by the event detector, the event data from the set of inputs comprising a first input of pilot input, a second input of data input from one or more aircraft systems that include Flight Management System (FMS) data, a third input of Air Traffic Control (ATC) clearance data, and a fourth input of log data generated by one or more checklists or pilot logs. ### Claim 13 (depends on 12) The method of claim 12, further comprising: configuring, by the processor, a hierarchy for determining a flight phase change or flight phase transition with the pilot input given a highest value. ### Claim 14 (depends on 13) The method of claim 13, further comprising: implementing, by the processor, a plurality of checks for determining whether data from the first, second, third, or fourth input triggers the determination by the event detector of the flight plan change or the flight transition for re-selecting of the current speech model in use by the transcription system. ### Claim 15 At least one non-transient computer-readable medium having instructions stored thereon that are configurable to cause at least one processor to perform a method for selection of a speech model in automatic speech recognition during runtime of a transcription system, the method comprising: determining, by the at least one processor, one of a number of flight events comprising flight plan changes and a flight phase transition based on event data received from a set of inputs; collating, by the at least one processor, a set of keywords associated with at least flight plan information for generating a wordlist in accordance with flight plan changes or the flight phase transition determined by the event detector, wherein the wordlist contains keywords associated the flight plan changes and the flight phase transition; determining, by the at least one processor based on the wordlist from the intelligent keyword generator, whether the wordlist is covered by a current speech model implemented in the automatic speech recognition of the transcription system; and in response to a determination that the wordlist is not covered by the current speech model, selecting, by the at least one processor by communication with a database storing one or more pre-built speech models, a pre-built speech model covering the wordlist. ### Claim 16 (depends on 15) The method of claim 15, further comprising: in response to the determination that the pre-built speech model covering the wordlist is not available, generating by the at least one processor, a new speech model for use as the current speech model for the automatic speech recognition of the transcription system for enabling coverage of the wordlist by the current speech model and for constraining the current speech model to at least the flight plan changes or the flight phase transition. ### Claim 17 (depends on 16) The method of claim 16, further comprising: determining coverage, by the at least one processor, of the wordlist by comparison of the wordlist to the keywords contained in each of the one or more pre-built speech models stored in the database. ### Claim 18 (depends on 17) The method of claim 17, further comprising: collating, by the at least one processor, keywords of the wordlist used in communication in the flight plan changes or the flight phase transition for comparison of keyword coverage in each of the one or more pre-built speech models stored in the database. ### Claim 19 (depends on 18) The method of claim 18, further comprising: receiving, by the at least one processor, the event data from the set of inputs comprising a first input of pilot input, a second input of data input from one or more aircraft systems that include Flight Management System (FMS) data, a third input of Air Traffic Control (ATC) clearance data, and a fourth input of log data generated by one or more checklists or pilot logs. ### Claim 20 (depends on 19) The method of claim 19, further comprising: implementing, by the at least one processor, a plurality of checks for determining whether data from the first, second, third, or fourth input triggers the determination by the event detector of the flight plan change or the flight transition for re-selecting of the current speech model in use by the transcription system.