Last month, Darian Johnson kicked off his crowdfunding campaign for Newt, a battery-powered, always-on, wall-mountable display that can give weather-related information, sync calendars, provide sports scores, display to-do lists, and do lots more.
The device syncs to Alexa via the Skill Messaging API. A user can press the “to-do” button on the display, which triggers the list skill, and instructs the skill to send my to-dos to the display. Johnson says the low power consumptions is what makes the Newt different: every component on the device’s motherboard was chosen with power consumption in mind. The device struck a chord with the audience on the crowdfunding site Crowd Supply. Johnson achieved over 400% of his funding goal in less than twenty-four hours.
Johnson is no stranger to Alexa. He has developed several skills that include Black History Facts, Blood Pressure Log and the Best Time of Day to Workout Recommendation. The latter uses machine learning to recommend the time of the day when you are most likely to maximize your workout intensity. Johnson also built the Mystic Mirror – a smart mirror that leverages Alexa Voice Service (AVS) to help people get organized as they are getting ready for work in the morning. The mirror won second place in Alexa's Internet of Voice challenge on Hackster.io.
Johnson says that building with Alexa is a no brainer for three reasons. The first relates to the rapid adoption of Alexa by consumers. He also points to the dynamic community of skill builders and device makers that continually engage and inspire each other – he finds the collaborative spirit reminiscent of the open-source ecosystem. Finally, he says that the tutorials and tools available to Alexa builders makes building with Alexa easy.
Johnson is particularly excited about building hardware devices with Alexa. He says that being able to develop both hardware and software gives him the ability to shape entirely new customer experiences. Also, as an engineer, Johnson enjoys working with the real-world physicality of objects that present an entirely different set of challenges from what one would encounter by exclusively focusing on the virtual world. He has chronicled his experiences building the Mystic Mirror and the Smart Home Lamp with Alexa on his blog.
In this article, he talks about the three steps hardware device makers can use to create Alexa-enabled devices.
Johnson says that the first step is to build a prototype – a preliminary model that informs future iterations of the device. Breadboard uses a breadboard to build a prototype.
“Think of a breadboard as a foundational platform with metal strips and holes,” says Johnson.
“You can then buy third-party components and use the breadboard to connect different modules and components to each other.”
Modules are discrete components that serve particular functions. For example, the power supply for the device could be housed in its own module. Johnson uses sites like Digikey and Mauser to source modules and components for the Newt.
“I took the different modules and components like screens and microcontrollers and connected them to each other,” says Johnson. “I then added, removed and moved components to make sure that I have a thorough understanding of how all these pieces would come together on the final device.”
The next step is to design designing a printed circuit board (PCB). PCBs help you provide reliable connections between different modules and components, and ensure that every component on your device is communicating with all other components and working as intended.
Johnson design his circuit boards by using a software like Eagle from Autodesk. He then conducts extensive research, and chooses a vendor to print his circuit boards. He has worked with a variety of sources in the United States of America and China.
“Designing a circuit board is like writing a novel. You have to iterate. My first design for the Newt didn’t function as expected. I had to go through four iterations of the design before I got to a place where things were functioning as expected.”
Depending on the device you are developing, Alexa offers a variety of options for device makers. Johnson used the Skill messaging API to allow users to send their to-do lists to Newt’s display.
He used AWS IoT Core to connect the Newt to the cloud. More specifically, Johnson used the MQTT messaging protocol. MQTT is a lightweight and widely adopted messaging protocol specifically designed for constrained devices – important for the Newt, given its focus on operating at a low power.
For the future, Johnson is excited at the prospect of leveraging Amazon’s recently announced support for Matter devices. He says developing Matter-compliant devices that have low power requirements will be a particularly interesting problem to solve.
“The starting point for me is to build devices that solve real-world problems,” says Johnson. “I am excited to keep building Alexa-enabled devices to help people like me: folk who don’t check for traffic delays till the last minute, don’t tune into the local news for the weather, are late to finding out about major news event, or forget something really important they are supposed to do on any given day.”