new feature is in development for Tower! Simulator that aims to significantly enhance immersion through the use of locally generated AI pilot voices. Titled “VOICES,” the upcoming system focuses on delivering realistic radio communications without relying on cloud-based services.
The VOICES system uses neural AI to generate pilot responses entirely offline, meaning no internet connection is required during use. According to the developers, this ensures that no data leaves the user’s PC, removing the need for subscriptions or external processing while maintaining consistent performance.
Each voice is trained using recordings from professional voice actors, with full consent, to provide more natural and human-like communication. The result is intended to replicate real-world pilot transmissions more closely than traditional text-to-speech systems. You can listen to the results now in the video above.
In addition to voice realism, the system also introduces dynamic audio environments. Each aircraft is said to feature its own cockpit sound profile, while radio quality will degrade based on distance. For example, an aircraft on final approach will sound clearer compared to one operating many miles away, reflecting real-world VHF radio behaviour.
This approach aims to create a more believable and reactive ATC environment within Tower! Simulator, particularly for users seeking a higher level of procedural realism during busy traffic scenarios.
No release date has been confirmed yet, but the feature is currently in active development.
