Challenges & Solutions
Unique Entities and Voices
Provide a SATCE solution with distinct speech from ATC and pilots on other aircraft.
During a training scenario, the customer’s communication environment is sterile: only students and instructors talk on the airways. To increase realism, the program needs a Simulated Air Traffic Control Environment (SATCE) that generates unique voices for every entity.
SERA fills the synthetic environment with automated human-interactive entities, including ATC and commercial aircraft flying around the airspace. Each entity sounds like a different person, supplying contextually accurate role-player dialogs to the environment.
Repeatable Diversions and Events
Create repeatable diversions and events during a flight, allowing students to target specific weaknesses.
SERA allows instructors to plan out scenario events and diversions (e.g., runway incursions) before the flight begins. If the student goes off course, SERA uses the latest artificial intelligence technology to react and correct the error, just like real ATC.
Updated Data Sets
Furnish accurate airport and navigation data that does not interfere with testing and certification.
Customers across programs share the benefit of new airports and data refinement. ASTi regularly updates SERA’s data package every 28 days, ensuring information is always correct. SERA also uses the same data packages as the Digital Aeronautical Flight Information File (DAFIF) and Aeronautical Radio, Incorporated (ARINC) 424 specifications. As a result, SERA’s updates do not interfere with the customer's testing and certification processes.
Nonintrusive Dialogue Prompts
Provide a developed speech technology that seamlessly understands and responds to students without getting stuck.
Many SATCE solutions suffer from faulty speech recognition. During flights, they get stuck in error loops that constantly prompt the student to repeat his or her response. Instead of a training enhancement, the technology becomes a frustrating distraction.
ASTi provides top-notch speech recognition and synthetic speech for training applications. We’ve conducted extensive development and testing to ensure that SERA’s speech recognition does not hinder training. SERA carefully listens to and interprets student responses, providing prompt and contextually accurate answers. As a result, students can focus on the training, not the technology.
Integrated Solution
Assimilate SERA into the customer’s existing simulator without significant development or configuration.
SERA works with a variety of simulators and can drive image generators in multiple formats, including Distributed Interactive Simulation (DIS) and Common Image Generator Interface (CIGI). The host system and instructor operator station can communicate with SERA’s public API. As a result, systems like traffic collision avoidance systems (TCAS) and Automatic Dependent Surveillance–Broadcast (ADS-B) can understand SERA entities. Administrators and instructors can easily configure and manage these settings in the SERA web interface. Contact ASTi to learn more about custom integration options.
Components Deployed
SERA
Simulated Environment for Realistic ATC
Features
- Speech recognition engine: delivers over 97 percent word accuracy for native English and English as a Second Language (ESL) speakers.
- Scenario generation: build and customize a training exercise that includes an ICAO flight plan and additional configuration settings.
- Flight plan generation: automatically generate the appropriate traffic and ATC interactions for any flight plan.
- Flight history transcripts: view past flights, capture data per student, review past student responses, and check student scores.
- Traffic plan generation: modify traffic density, call signs, departure and destination airports, and the type of aircraft that appears at an airport.
- Speech model and pronunciation tuning: SERA's speech recognition models are highly tuned to support over 30 ESL and native English accents.