Whether it’s Siri, Google Assistant, Alexa, or Cortana, there is a very good chance that you have a voice assistant within earshot of you right now. User interactivity is evolving from a pure physical interaction model to one that is driven wirelessly through voice. Customers are progressively expecting the devices that they buy to be connected and capable of integrating with their favorite flavor of voice assistant.
In this blog, I’ll talk more specifically about the options available for integrating with market leaders like Amazon Alexa and Google Home voice assistants, to help you understand the interaction model that makes most sense for your business and device types.
Alexa and Google Home Voice Assistant Integration Comparisons
Amazon and Google have architected their integration paths using very similar patterns, which is beneficial to integrators and manufacturers of Internet connected devices. The underlying solutions are different for each, but they speak the same vocabulary in many ways allowing for us to evaluate both integrations in the same breath.
The Options: Smart Home Skill versus Custom Skills
The way Amazon and Google have arranged the integrations boils down to two options – Smart Home Skills and Custom Skills. I’m using Amazon’s vocabulary here, but it is synonymous with what Google offers – Smart Home Action and Custom Action.
The primary difference between the two:
- Smart Home Skills follow a very prescribed interaction formula for a limited set of device types directed by Amazon or Google
- Custom Skills leave the interaction model and device types up to the implementer’s discretion
Let’s jump into each option in a little bit more detail.
A Custom Skill/Action is the most common type of interaction with Amazon Alexa and Google Home. As the name implies, Custom Skill/Action allows for the most customization when interacting with the voice assistants – reading news, playing a game of Jeopardy, or getting the latest scores for your favorite sports teams.
To implement this kind of Skill, the integration team will be responsible for uniquely naming their Skill, creating a voice user interface, and implementing the backend services to perform the requests. The name of the Skill is directly used for invocation, for example, if you named your Skill “Thermostat 3000,” a user would be required to say something like: “Alexa, ask Thermostat 3000 to set the temperature to 74 degrees.”
The advantages of this approach are straight forward, you are granted the flexibility to do anything that you can technically implement. However, this flexibility comes at the cost of a longer time-to-market because you will need to implement every detail of the interaction.
You will be responsible for:
- Creating the sample phrases that a user can invoke
- Modelling the error responses that may come from those interactions
- Translating those interactions to other languages if required
- Developing backend components to handle the requests for your devices
Smart Home Skill/Action
A Smart Home Skill/Action provides a special interface and integration with Alexa and Google Home voice assistance for certain device types typically found in the emerging Smart Home space.
Supported device types are:
- Door locks
- Entertainment devices like TVs and receivers
- Smart home cameras
- Thermostats and fans
- Microwave ovens
Interaction with the devices has been condensed to a common set of patterns and provided to implementers to utilize, if your product fits within those patterns. Unlike Custom Skills, you do not need to name your Skill and the invocation of the skills is much more streamlined with the voice assistance. For example, if a user wanted to raise the temperature using your thermostat Smart Home Skill, the user would simply say – “Alexa, set the temperature to 74 degrees”.
Smart Home Skills still require the implementation team to build the backend services to interact with your device, but Amazon and Google provide all the voice interaction and language localization features. By providing the voice interaction models, the time to market will be much shorter than a Custom Skill, but you will be constrained to only being able to utilize the predefined voice interactions. You cannot modify or change the interaction model. If you have a special use case in your product, you might need to implement that as a separate custom skill if it is a “must have” feature.
Another benefit of Smart Home Skills is the potential opportunity for co-branding by being able to go through a certification process with Amazon. Passing the certification process with Amazon allows for the products to have branding on their packaging like “Works with Alexa” and gets featured in a special sub-store of Amazon.com for “Works with Alexa” products.
In general, our recommendation at eLogic is to utilize Smart Home Skills if your device types align with what Amazon and Google provide.
- Reduced time to market
- Free language localization
- Branding opportunities, which can be very compelling to any business entering this space.
If you don’t align with the device types Amazon and Google have implemented, or require specific features not found in their offerings, a Custom Skill will be the best option for integration.
Azure as Foundation
Microsoft Azure services can provide the foundation for voice assistant capabilities, and eLogic has the expertise to help our partners bring their products to these powerful voice platforms. We handle the full stack of development in this space, including developing the cloud microservices to communicate with the Voice Assistant platforms, deploying an IoT Hub to handle the device management, and developing the software necessary to make devices smart and connected.
If you’d like to learn more, get in touch with us -- we can help you connect your devices to the rapidly expanding world of voice enabled smart assistants.