We already have voice assistants on our devices so why do we need another home device? The rationale behind each of the companies in this space is to fortify their businesses. Amazon does not have a presence in the mobile phone market so it has developed the Echo to get into the home – mainly to drive sales to its Amazon store. Currently Amazon is the market leader, with Google coming in second place.
The advantage for Microsoft is that Cortana is used by millions of Windows users already and that is now available in the home. Cortana is also a powerful digital assistant. Siri is in a similar position, so it would make sense for Apple to leverage Siri to the smart home speaker market. Apple could also bring Apple TV into the picture.
However, a key issue with this nascent market is that none of these digital assistants can talk to each other or co-operate in any meaningful way. Further, this is not an easy thing to do as the intelligence behind these systems is in the cloud. At the moment, having two different systems in the home is potentially a problem.
Amazon has taken a step further by partnering with NXP to create the first complete NXP reference platform for Amazon Alexa with far-field technology to simplify development of high performance voice-enabled devices.
Designed to simplify the creation of new voice-control devices, the NXP reference platform for Amazon Alexa on i.MX is a complete system containing Amazon’s 7-microphone array design, far-field audio processing technology, and AVS to enable customers and partners to create their own high performance devices with Alexa.
NXP’s i.MX applications processors offer the scalability to meet the requirements that are unique to each customer design within the Internet of Things (IoT). The applications processor portfolio features the i.MX 6 series for general purpose applications, i.MX 7 series for low power applications, and the new advanced i.MX 8 series to enable rich interactive high performance multimedia and audio experiences.
In effect, the partnership has enabled anyone to put the Alexa voice service into their products, enabling Amazon to grow the ecosystem. The reference design also ensures that devices associated with Alexa benefit from high quality of the far field algorithms Alexa uses.
One of the reasons Amazon chose NXP according to Leonardo Azevedo, Director of i.MX Applications Processors for Consumer Market at NXP was the fact that NXP cover the widest range of customers directly and through distribution, which fits Amazon’s goal of addressing as many developers as possible – in order to expand the ecosystem as much as possible.
Today machines have reached a breakthrough in understanding commands. Even humans are not 100 percent accurate, but machines are now getting close to human accuracy from a low level just a decade or two ago. One of the reasons for the success of Alexa is that it has become generally accepted today that machines can recognise voice with enough reliability to be useful. In addition, people have got used to digital voice assistants like Alexa and Siri.
For Amazon the next step is enable voice technology without the need for Echo – allowing other manufacturers to put the Amazon voice technology into a variety devices in the house such as coffee machines, fridges, lights and even in cars, and have all this connected to the ecosystem, adds Leonardo Azevedo.
This is where the reference design comes in – allowing manufacturers to adopt these high performance voice algorithms faster. The focus in the future will be on AI, machine learning and the smart home.
“We believe the ecosystem will be more useful with other units connected in the house,” says Leonardo.
The goal for Amazon is to make it easier for consumers to shop on their platform and drive their Amazon Prime platform. With others pushing their ecosystems into the house in various forms, Amazon wants to have the first mover advantage.
Leonardo adds, “ Voice will be huge in the house and once all devices have voice – everything will work together.”
One caveat here is that there are no common standards and not a lot of effort is going into solving the problem. It is possible that this problem might be resolved by Christmas this year says Leonardo.
It is reasonable to expect a big push towards voice activating devices as this is a natural interface for humans. However, for such systems to work in mobile devices such as cars the need for a fast connection to the various clouds that enable the voice systems such as Siri, Alexa and Google Assistant is required. Further, the interests of the consumer, these systems will also need to co-operate to a certain degree with each other to be adopted en mass.
To conclude, voice might be the next big thing in ushering in the convergence of mobile connectivity and cloud services to consumers from entertainment to online shopping and even healthcare services.