Intelligent agents for automation

The sophisticated technology within Phonak Marvel allows it to function as a wearer’s personal intelligent agent — providing confidence that devices will function optimally and automatically.

Don’t we all prefer things to be automated these days? Intelligent driving with cruise control, distance regulation and lane centering, automatic climate control in the car and at home, quick-sync connection of all our paired devices, voice commands, and so forth…. “Time is money” and “Convenience is king” go the sayings, and these days we seem to pack so much into our daily programs and technology is expected to keep pace.

This is no less relevant when it comes to hearing and participating in life. For some years now, hearing devices have been classifying sound in the various environments frequented by wearers in order to automatically adjust the settings to provide the most optimal listening experience. In fact, Phonak technology has been doing this since 1999 with the first iteration of sound classification, AutoSelect, in the Claro platform. Over the years, auto-classification has developed to include more sound scenes into the system training for added robustness and accuracy.

Marvel-ous automation

Recently, Phonak released Audéo™ Marvel, a multifunctional hearing aid delivering a ‘Love at first sound’ experience. Marvel has been enhanced for spontaneous first fit acceptance, meaning that new wearers experience exceptional sound quality1 from the first very first moment. Streamed media inputs are now included in the auto-classification system, AutoSense OS 3.0, in the same way in which acoustic inputs are. This means that media signals are assessed for the presence of music vs ‘non-music’ (or speech), and settings are automatically adjusted accordingly for top-rated streamed sound quality2. Dual path compression is available and activated where required (i.e. for speech in noise and media streaming), in order to maintain the natural modulations of speech. Vent loss gain compensation, providing low frequency gain of up to 35dB, is also applied during streaming of audio signals, for a full and rich sound, without compromising inputs received directly by the hearing aid microphones.3

The Phonak signature Binaural Voice Stream Technology™ makes its comeback in Marvel, allowing for the full audio bandwidth to be streamed binaurally and in real time, improving speech understanding and reducing listening effort in challenging listening situations.4 The Binaural Voice Stream Technology allows functionality such as Phonak StereoZoom – binaural directional microphone technology. AutoSense OS 3.0 automatically switches into StereoZoom to create a narrow beam in especially challenging listening situations, and recent studies have shown that this results in reduced listening and memory effort,5 leading to increased participation of wearers in conversations in noise.6 That’s quite a package!

Connecting humans and intelligent agents

Language and communication technologies play a key role in enabling humans to interact with intelligent agents in a natural and intuitive way. The sophistication of logic and reasoning technology within Phonak Marvel allows it to function as a wearer’s personal intelligent agent. The training phase in the development of AutoSense OS 3.0 is comprehensive including numerous ITTT (If This, Then That) algorithms, allowing the system to learn the most appropriate settings in a broad range of acoustical environments.

It is able to blend different settings, tailored to the wearer’s listening needs at any time. Providing confidence that devices will function optimally and automatically is key to enabling wearer’s to participate fully in life, without limitations.


We invite you to read Tania’s previous blog post on how hearing care professionals rated Phonak Marvel at EUHA 2018.



1 Jansen, S. & Woodward, J. (2018). Love at first sound: The new Phonak precalculation. Phonak Insight. Retrieved from

2 Legarth, S.V., Latzel, M. & Rodrigues, T. (2018). Media streaming: The sound quality preferred by hearing aid users. Field Study News. Retrieved from

3 Woodward, J., Pislak, S. & Kühnel, V. (2018). New dual-path processing of vent loss compensation enables a clear and rich sound quality when streaming. Phonak Insight. Retrieved from

4 Winneke, A., De Vos, M., Wagener, K., Latzel, M., Derleth, P., Appell, J., & Wallhoff, F. (2016). Reduction of listening effort with binaural algorithms in hearing aids: an EEG study. In Conference of the American Auditory Society. Arizona. Retrieved from

5 Winneke, A., Latzel, M., & Appleton-Huber, J. (2018). Less listening – and memory effort in noisy situations with StereoZoom. Field Study News. Retrieved from

6 Schulte, M., Meis, M., Krüger, M., Latzel, M. & Appleton-Huber, J. (2018). Significant increase in the amount of social interaction when using StereoZoom. Field Study News. Retrieved from