Many mobile applications use sensors to monitor various activities- movement, sound, temperature, and the like. The sensors, if “always on” in full power would shorten battery life. Recent chip developments enable applications and devices to operate in a low power mode until they detect the signal that wakes them up and activates them for the next task. The most common application is to use a motion sensing chip to wake up the device. Some applications use voice commands to “wake” the device from the low power detection mode and switch to the “recognition” mode, applying their algorithms to spoken words and phrases. This is ideal for many wearable devices as well as other battery operated devices.
Since devices utilize far less power, devices use smaller batteries and are smaller and lighter.
Devices that utilize voice activated wake up can be found in both the home and in industry. As the IoT sector expands, expect to see more voice activated applications that enable users to interact with a device completely hands free.
Voice Activated Wake Up Devices
Some applications currently on the market that use voice activated wake up include:
- Amazon Echo: With the Amazon Echo, the user addresses voice commands to Amazon’s Alexa to play music or control other smart devices in the home. A broad range of compatible apps are available and are being added.
- Xbox Kinect Adapter and Sensor: The Adapter and Sensor connect to Xbox One X/S gaming system to allow the user to control games by motion and spoken word. Allows you to control the entire system with verbal commands or motion. Xbox is a Microsoft product.
- aTTo: The voice activated recorder has 20 hours of battery life and 90 hours recordings capacity in the size of a thumb drive.
- Pinron: The voice controlled toy car recognizes go forward, go back, turbo (accelerate), stop, and spin commands.
Chips That Enable Voice Activated Applications and Devices
|Manufacturer||Chip||Features||Power during listen mode|
|Vesper||VM1010 Wake-on-Sound Piezoelectric MEMS Microphone||9 uW|
|ON Semiconductor||ON Semiconductor BelaSigna R281||Voice activation in portable devices, acting as an ultra-low-power, always-listening voice trigger.||< 300uW|
|DSP Group||DHX91 SoC||A low-power chipset solution for home automation and security. It has audio capabilities and a powerful ARM926™ processor.||3 uW|
|Knowles||Audience eS700||Voice wake software which enables the device to be in always listening mode, ready to understand and act upon verbal commands, without the need for interaction with touch, without significantly impacting battery life. Natural language and configurable voice commands.|
|Synaptics||CX20924 CX20921||AudioSmart™ Far-Field Voice DSPs. Low-power WoV mode with embedded third-party voice wake-up engine. 2 or 4 microphones.||70 mW|
Development Platforms and Solutions
|Platform or Solution||Partners||Processors||Use Cases|
|VM1010 Dev Platform||Vesper, DSP and Sensory||Vesper VM1010 DSP Group DBMD6 – an ultra-low-power, always-on voice and audio processor||Initiate voice processing through sound detection. See https://vespermems.com/|
|AudioSmart™ Development Kits for Amazon AWS||Synaptics and Amazon||Synaptics CX20924 or CX20921||Help device makers quickly turn new ideas into marketable Alexa products that offer an ideal voice user experience.|
|Google Assistant SDK||https://developers.google.com/assistant/sdk/overview||The Google Assistant SDK lets you add hotword detection, voice control, natural language understanding and Google’s smarts to your devices.|
|SiriKit||Apple||https://developer.apple.com/sirikit/||SiriKit enables your iOS apps and watchOS apps to work with Siri, so users can get things done using just their voice. Your content and services can be used in new scenarios including access from the lock screen and hands-free use. Future plans include expansion to Apple HomePod.|
|Alexa Skills Kit||Amazon||https://developer.amazon.com/alexa-skills-kit||Allows you to add Alexa Voice Service (AVS) to add intelligent voice control to any connected product that has a microphone and speaker.|