The Smart Home is Listening ’” and Talking

• Features

• 19/04/2016

An increasing number of consumer products now have voice command capabilities, ranging from streaming media players to smart home products. Products like Apple’s Siri and Amazon Echo have captured the attention of consumers and are helping to drive consumer interest in chatting with devices. In addition, advances in understanding natural language are opening doors for automation, information access, and hands-free control.

The automotive, gaming, streaming media devices, and pay TV industries are today’s main players in voice command systems. Audio capabilities provide alternative remote control options for GPS systems, phone calling, games, music selection, and TV programming. Using voice allows an easier adjustment of previously created scenes than other control methods and places more control in the hands of the user rather than the system. Audio sensors detect smoke alarms, glass breaking, or unusual sounds and send alerts to homeowner and security monitoring services.

Amazon Echo uses natural language to handle a number of pre-programmed retrieval tasks (e.g., alarm, lists, weather, music, traffic) as well as to answer questions based upon Wikipedia. IBM is offering Watson technology, which can understand natural language questions and search for information; Watson holds the potential to learn over time. Apple’s Siri and Google Now both use natural language to process information requests on mobile devices.

Some production homebuilders have begun offering smart home systems to their customers—and some of these systems include (or will soon include) audio solutions. Lennar now offers the Nexia Home Intelligence line to its new home owners; Nexia announced a natural language voice control pilot with Vognition in early 2015. Some homebuilders are expected to also embrace audio analytic solutions for their development homes as a means to create differentiation. People normally prefer a natural language experience and voice command offers more reliable control relative to the users’ intent with minimal physical exertion. Aging baby boomers want to maintain their lifestyles; in some cases this will demand they have products with effortless control and alerts. Those needs can be a powerful force in this space. Their desire to age in place will drive adoption of audio solutions. Voice reminders, voice alerts and voice recognition are all important audio elements.

Smart home control prices are decreasing drastically; prices for audio solutions are no exception. Over time, a commoditization of home automation solutions will significantly reduce the cost of the required solution to achieve a “smart home.” Affordability will reduce the perceived risk of trial. At the same time costs decline, product and technology will improve, making the idea of using voice more appealing to people who like the idea — even if no specific physical factor creates a need for voice control.

Natural language knowledge navigators are steadily improving and voice analytics will continue to improve dramatically.  There will be no reason not to include audio capability in smart home solutions. Pattern recognition technology is getting to a level of recognition and accuracy that makes the interface more natural. As the technology is catching up, use of audio interfaces in the automotive and mobile phone space is making users more comfortable with voice control. Cloud-based data analysis has increased available computing power, allowing the use of audio analysis principles that have been around for years.

Success Factors
In order for voice controls to become common throughout the smart home, several key areas must be addressed. Security is critical. Access by voice control has to be limited to specific individuals for applications involving security to and within the home. A voice-controlled smart system needs to be controllable only by the homeowner and his or her designated persons. If there is an intruder, traditional security systems take over while the audio control system goes into auto mode and locks out commands. To accomplish this, an audio control system identifies the person giving the commands, either by code or by vocal pattern match. Few current audio analysis systems have this level of sophistication.

At present, microphones for voice control are bulky and expensive. In order to have whole home control capability, audio capture must occur in each room. Standalone microphones cost $150-$200 and computer capability costs over $150, driving the cost to roughly $350-$400 per room. The industry will drive this cost down considerably over time. Rather than standalone microphones, it is possible that devices could incorporate small microphones that would enable whole-home control capability.

Today, many voice solutions are inadequate, and consumers all too often are disappointed and angry at the product’s inability to meet expectations. In some cases, voice solutions force the homeowner to change how he or she normally interacts by voice rather than allowing natural voice tones. For people hesitant to use voice as a control mechanism, a sort of stage fright occurs that causes voice tone changes. Experience has shown that the more naturally users can or are allowed to speak, the better the user experience.

Low awareness or lack of appreciation of the value of voice control ultimately inhibits adoption. In order for audio analytic solutions to gain greater traction, companies will have to communicate to the consumer the value of using voice as a control mechanism in application specific ways.

Once product and software companies realize the differentiation that voice control can provide, and believe in that, they will develop and market products offering this control as an option. With more products available and with specific control applications, homeowners can experience the capability of audio control. Assuming good experiences in demos or by observing applications in the homes of friends and family, more adoption will occur. The challenge is to determine the useful services that can be enhanced by audio control.

Ideally, voice would be one of a number of device communication methods, including touch and gesture, enabling consumers to communicate with devices as naturally as they communicate with other people.

Brett Sappington is a senior director of research at Parks Associates. He leads the firm’s services research team, which publishes industry insights on access and entertainment services, digital media, OTT, cloud media, video gaming, and technical support services. 

Barbara Kraus is a director of research at Parks Associates, where she studies the connected consumer electronics market. Barbara’s expertise is in delivering and monetizing consumer insights through the provision of compelling product and service technologies and experiences.

Park Associates