DW Embedded

Offline Voice Control Modul

Welcome to the website of DW Embedded. We are OEM manufacturers for B2B offline voice control modules. Our solution is mainly used as an offline voice control assistant for supporting operations in medical applications and is now finding its way into applications in the broad market such as Industry 4.0, Smart Home, Robotics, Home Entertainment and much more. The main features of our voice control system are the high quality, offline and the technically most affordable hardware solution. With our solution we revolutionize the world of offline voice control.

make devices speaking with you


Low recognition rate

Small distance (30cm)


Teach-in required




High recogntion rate

The module offers the highest detection rates even in demanding applications such as echo and background noise in the operating room.


More than 14 languages are available for the module factory-fitted. 


The modules are very compact despite completely self-sufficient functionality and can be integrated into almost any PCB. With long-term availability.

Far-field up to 16m

Up to a range of 16m, commands with a high recognition rate can be detected and processed.

Speaker independent

No teach-in necessary.

100% Offline functionality

The modules are completely self-sufficient and do not require an Internet connection to develop their full functionality.

Hybrid Function

1. Voice Control Module

The voice control module is very compact, the microphones can be positioned according to the customer’s wishes, is optimized at maximum cost, the direction can be fixed via a beam width and contains the following speech control functions. 

A speaker independent wake-up word is used to activate the voice control module.

A personalized wake-up word can be trained and is used to activate the voice control module.

A secure wake-up word can be trained and  is used to activate the voice control module. For example to get acess to an house door.

Natural speech recognizes the context in an unlimited number of different sentences. For example, turn on the light, it’s too dark here, please make it brighter here, turn on the lamp, etc. Here the possibilities of variations are unlimited.

After the wake-up word commands are detected and then can be controlled via an interface USART, SPI, GPIO, etc. and / or voice outputs are performed. 

The offline assistant can store and play audio data. Then the module is able to answer.

2. Audio Front End

The module also work as a hybrid function, so no voice control solution from DW Embedded needs to be used, only the prefiltered voice signal can be used via USB or I2S.


Example DWe voice control modul:

voice control IOT

About us


We burn for speech control systems. As a highly driven and ambitious team of young professionals, we offer voice control solutions and audio front end solutions for high end speech control applications. Our focus is on the active operation of the medical field. Due to the high requirements in the medical field, the speech recognition system can also be used with very good quality in areas such as smart home, automotive and in industry.


In the current market situation there is a trend for the implementation of the speech recogntion, which will now become more and more established in the market and shape our future. We give impulses and stay curious, so we can certainly inspire you. Our goal is to shape the future with you and to implement the speech recogntion in your products.



Do I need a host processor? No, you don’t need a host processor, there are I/O’s, SPI, UART and more on request.

Can the system be personalized to a specific voice? Yes, the language can be personalized, it already works and will soon be implemented in the module.

How many languages can be stored? Currently there could be stored about 20 languages, on request also up to 60 languages. 

Does the system need to be trained? No, the system does not need to be trained.

Can words be added later? Yes, Words can be added later.