Weighing the trade-offs in smart speaker designs


Many home appliances are adding voice-recognition capabilities to enhance the user experience. But the one product that ties the whole ecosystem together is the smart speaker.

In addition to playing audio content via Wi-Fi®- or Bluetooth®-enabled devices, today’s smart speakers can communicate with other appliances, such as lights, door locks or smart thermostats, as illustrated in Figure 1.

Figure 1: As a media player, smart speakers must be simple and elegant with quality sound. As a smart home hub, they must provide accurate voice recognition and connectivity to the entire suite of smart devices in the household.

With that much functionality, a lot goes into the design of a smart speaker, and many of the design decisions involve trade-offs in terms of performance and efficiency. The two main design features we will discuss in this post are the audio and the human-machine interface (HMI).

 


Read more about smart speaker solutions

Our white paper weighs the key trade-offs to consider in your smart speaker design. 

On the audio side, the first thing to consider is accurate voice recognition. The defining feature of a smart speaker is that it can listen to and respond to voice commands from users. Typically, there are two types of microphones used for this feature: digital and analog. The main difference between the two types is whether the analog-to-digital converter (ADC) is integrated or external to the microphone. A digital microphone integrates the ADC, which requires a smaller ADC that causes a higher error rate for voice recognition. With an analog microphone, the ADC is external and provides performance benefits, leading to more accurate voice recognition.

On the other end of the signal chain is the audio output, which is just as important. By using speaker protection as well as the built-in digital signal processor, the speaker can be small, efficient and provide premium sound quality in a small form factor 

Another decision to make is what type of HMI to offer. Along with the audio response, smart speakers typically also include a visual response such as an LED ring to show off a pattern indicating status. With this, there are design considerations in terms of processing power and overall system power. TI’s LED engine enables a lot of the processing power to be offloaded onto the LED engine, enabling brilliant color and pattern generation.

There are other considerations such as power, battery management, input interface, display technology and other techniques, all being incorporated into smart speaker designs. I recently co-authored a white paper that further discusses these smart speaker design features.

As the market grows, smart speaker functionality will start to appear in other appliances such as sound bars, TVs, set-top boxes and many other products, bringing new opportunities for designers to differentiate their designs.

Additional resources