Local voice systems enable efficient smart home control

Local voice systems enable efficient smart home control
Starting last year, voice control has been an important issue in the smart home space. People tend of be more familiar with Amazon Echo or Google Home when it comes to using voice to control home devices. The way it works is that upon receiving a voice command, a smart speaker will pass the voice to the cloud for processing before a control signal is passed to a home device.

There is a different way of doing it, that is, local voice control, which does not require passing the command to the cloud via Wi-Fi. There are many advantages of using a local voice control system. It features low latency, and runs on a more stable and secure internal network.

In a local network environment, a voice command travels from a microphone to a gateway which then passes the command to home devices. Because the signal does not travel to the Internet, lower latency can be ensured. Also, users do not need to worry about unstable Internet connection, because the signal is only passed around in the house. Since home devices are not linked to the Internet, hackers won’t have chances to control home devices and security can be ensured.
 
James Huang, CEO of
Philio Technology

The merits of local voice control are acknowledged by many smart home system providers. One of them is Taiwan-based Philio Technology Corp. Controlling every home device from the cloud sounds more complex than it needs to be, according to James Huang, the company’s CEO. “A simple task like opening curtains at home should not require processing in the cloud before sending a command back to the device.”

Drawback of smart speakers

Another advantage of using a local voice system is that it does not require the purchase of an expensive smart speaker.

Smart speakers from big companies have great audio systems, which are usually more than sufficient for voice control scenarios. Also, a smart speaker may have up to 6-8 microphones embedded so that it can receive voices from different directions and for noise cancellation purpose. These functionalities, however, will add up to costs. As such, it may not be feasible for a user to place a smart speaker in every room, which is the ideal setting for voice control.

Chen Wen-bin, Manager
of the IoT Business
Section at Wifly-City

Users should be able to access voice control anywhere at home. “Voice control should not be constrained in a certain area. When in the bedroom, I don’t want to go to the speaker placed in the living room to turn off the light,” said Chen Wen-bin, Manager of the IoT Business Department at Wifly-City, a company that provides smart home solutions for the retail market.

Alternative voice control

Local companies that enable smart homes are rolling out alternatives to smart speakers. Philio, for example, plans to introduce a battery-powered speaker that has the shape and size of a baseball. The speaker may be embedded in the wall in the living room, kitchen, bedroom, etc., thus allowing a user to give voice command more conveniently. The product could be a good replacement to a smart speaker, the placement of which is constrained by the power line it connects to.

Electronics manufacturer Good Way Technology has co-developed a smart panel which can receive and pass voice commands to other devices. The panel is embedded with a voice IC that performs voice recognition and processing. 20-30 voice commands are stored in the panel.

A portable voice control device is more common. Unless being tapped or triggered by the wake-up word, the device is in the sleep mode to save battery power. Wifly-City has such device while Good Way has plans to develop and launch one in the near future.

All these voice-receiving devices work by passing voice commands to a gateway that relay signals to home devices. A gateway essentially can talk to almost every kind of home electronics. The chosen communication protocols, however, separate companies making gateways. Common standards Z-Wave, Zigbee, ULE and RF have all been used.

IoT company Full Enterprise has developed local voice control by inserting control modules into devices. The company’s voice control may turn on one single device or multiple devices to enable a scene. In the future, Full Enterprise also plans to develop voice control on a smartphone to control home devices.

“While voice control is convenient for all age groups, it is especially suitable for the elderly or the disabled,” Full Enterprise’s General Manager James Yih noted.

Challenges in making voice devices

Tali Chen, CVP
Corporate Development
& Chief Evangelist at DSP
“One of the main challenges of making voice-enabled devices is that voice can hardly be detected if the source is three meters away. Also, engineers need to figure out how to factor out background noises,” said Coco Chen, Deputy Marketing Director at China-based Leedarson Lighting, which has enabled voice control for a variety of home devices, including lighting, curtains and TVs, made by Leedarson or other brands. Leedarson’s team of engineers have developed the needed computation power for noise cancellation and also tailor-made microphone designs for best voice-receiving effect, Chen added.

DSP Group, which makes chips for a variety of communication products, believes that ULE is the better protocol choice. Tali Chen, CVP Corporate Development & Chief Evangelist at DSP, said that ULE is based on the DECT technology that is used in cordless phones and therefore has better voice transmission capability. ULE also uses an interference-free spectrum. “As such, ULE has much, much better transmission range in both indoor and outdoor use cases, compared with other protocols like Wi-Fi, Z-Wave and Zigbee,” Chen said.

Limitations of local voice control

CD Liu, Director of Good
Way Technology’s SIoT
Research and
Development Division
Local voice control system has its drawbacks too. Compared with the cloud, the local voice IC does not have the voice processing power similar to that of the cloud, said CD Liu, Director of Good Way Technology’s SIoT Research and Development Division. The cloud is able to distinguish different accents, background noise and more accurately understand spoken words. It also has a much larger storage space, allowing for unlimited number of command storage.

Nonetheless, local processing and cloud processing can co-exist. In fact, many companies have created products that can connect to the cloud as well as products that operate in a local network.
Share to:
Comments ( 0 )