Audio technology developed by the Fraunhofer IDMT in Oldenburg used in Deutsche Telekom’s smart speaker

News /

The Fraunhofer IDMT in Oldenburg has developed the audio technology for Deutsche Telekom’s smart speaker. At the heart of the work is the optimized interaction of speakers and microphones for voice control even in a noisy environment. What’s more, the experts have achieved particularly good acoustics for telephony and music playback in a small enclosure. The product will be launched on 31.08.2018 at IFA in Berlin, the trade exhibition for consumer electronics and home appliances.

© Deutsche Telekom
With their audio development of Deutsche Telekom’s new smart speaker, audio experts from the Fraunhofer Institute for Digital Media Technology IDMT, Branch Hearing, Speech and Audio Technology HSA in Oldenburg have succeeded in creating a superb sound in a small space.
© Fraunhofer IDMT/Markus Hibbeler
Dr. Jens-E. Appell, Abteilungsleiter Institutsteil Hör-, Sprach- und Audiotechnologie am Fraunhofer IDMT-HSA in Oldenburg
Jan Wellmann, Gruppenleiter für Audiosystemtechnologie am Fraunhofer IDMT-HSA in Oldenburg

Oldenburg, 31.08.2018. With their audio development of Deutsche Telekom’s new smart speaker, audio experts from the Fraunhofer Institute for Digital Media Technology IDMT, Branch Hearing, Speech and Audio Technology HSA in Oldenburg have succeeded in creating a superb sound in a small space. By refining and combining microphone and loudspeaker technologies, they have achieved a very robust voice control system that also works reliably when listening to music at the same time and in a noisy environment.

What was special for the product developers in Oldenburg was the fact that they were involved in the project from the outset and able to contribute their know-how right up to end-of-line testing. This allowed a close and trusting relationship to grow with Deutsche Telekom at a high technical level. “Our experience shows that our clients often want an integrated system and increasingly also trustworthy support in every stage of implementation. From concept development to first hardware and software prototypes to the start of production – we were able to map all this out for Telekom in the way they wanted,” says Dr. Jens-E. Appell, Head of Branch Speech and Audio Technology HSA at the Fraunhofer IDMT in Oldenburg, underlining the good cooperation between the two organizations.

“Through controllable directional microphones and signal enhancement for recognition of voice commands, we were able to give the device superb ‘hearing’.  This also makes itself noticeable when using the integrated hands-free function to make phone calls as well as when accessing other services and operating entertainment equipment,” adds Jan Wellmann, Head of the Audio System Technology Group at Fraunhofer IDMT-HSA in Oldenburg.

“Since the quality of our products is always in the foreground, the main focus for our smart speaker was the acoustics. In first discussions, we quickly recognized that the experts in Oldenburg can assist us throughout the entire development process. Starting with the finetuning of our first draft designs, followed by the many questions that arose and each stage in product optimization during the course of the development work to quality assurance during the final product inspection, the Fraunhofer IDMT was a reliable partner that accompanied us along the way,” highlights Dirk A. Böttger, Vice President Product Management Fixed Line Devices at Deutsche Telekom.

And: “We believe that voice control in intelligent products has come to stay. This means that it’s time for our smart speaker which complies with German standards. At the same time, the system is explicitly designed as an open platform onto which each and every company and ecosystem can dock – from large supermarket chains such as REWE to corner shops,” continues Böttger.

Numerous functions in compliance with European data protection rules

With the personal assistant, voice control can be used to operate connected devices in the home such as TV and Telekom’s smart home applications – without the need for a remote control. The personal assistant can also be used to make telephone calls – the audio quality is excellent. Other applications are available via cooperation partners such as Radioplayer and the news programme “Tagesschau”. Users of Amazon’s cloud-based Alexa voice service can also access a large number of apps with Deutsche Telekom’s smart speaker. The new voice-controlled assistant conforms to the highest data security requirements. Data processing takes place exclusively within the European Union and complies with Telekom’s high data protection standards.

 

About the Branch Hearing, Speech and Audio Technology at the Fraunhofer Institute for Digital Media Technology IDMT in Oldenburg

The objective of the Branch Hearing, Speech and Audio Technology of the Fraunhofer IDMT is to transpose scientific findings related to hearing perception into technological applications. Its applied research priorities are the enhancement of sound and speech intelligibility, personalized audio reproduction, and acoustic speech and event detection. Through scientific partnerships, it has close links with the Carl von Ossietzky University, Jade University and other institutions in Oldenburg engaged in hearing research. It is also a partner in the »Hearing4all« Cluster of Excellence.

WE MAKE IT LISTEN! "We teach your devices to hear."

With the speech recognizers developed at the Fraunhofer IDMT, enterprises can operate their machines reliably even in a loud environment and, if wanted, without transferring data to the internet. Or they can incorporate the speech recognizers in their own products and in this way create unique selling points as well as generate new value added for their customers. In the area of industry 4.0, amongst others, the Fraunhofer IDMT is using acoustic event monitoring processes to detect conspicuous noises in machines or in the context of product quality control. In this way, damage can be avoided and downtimes kept to a minimum.

Last modified: