Most of the coverage from Amazon’s devices event yesterday focused on the $ 249.99 Ring camera, which is attached to a drone and will fly around your house when an alarm is triggered, looking for invaders. It was — and will largely continue to be — about the expansion of Amazon’s surveillance network into your home, the actual need for this sort of device, and likely how this device is actually pretty cool because it makes it easy for a normal person to program the flight path of a drone. That’s some impressive AI!
But what I saw at the event wasn’t the devices; it was the data. In the almost six years since the Amazon Echo first launched, Amazon has been seeding this hardware inside our homes, building out an understanding of exactly what we consumers do in our homes and how we do it. This year’s device event showed just how well Amazon has used that data not just to launch new devices, but to place itself as an essential translator between what consumers want and what developers’ devices should deliver.
In short, Amazon is behaving exactly like Amazon. Look at its history with Amazon Web Services, where it offers cloud computing, storage, data analytics, and an ever-increasing array of tools for developers. It uses the data from companies that build their own developer tools on Amazon’s cloud, or the developers who link these tools to the AWS cloud to monitor their own services, to understand how various services fare before then launching its own competing product. In the retail market, meanwhile, it analyzes what people buy in order to decide on the types of products it will make under its own brands.
So in the smart home, it stands to reason that Amazon is doing something similar. In the case of devices, Amazon can be blatantly obvious, launching products that directly compete with companies it has invested in or met with to discuss Alexa integrations. But it doesn’t need to meet with a company to know which way the market is heading. All it needs to do is see what devices are getting tied into the millions of Echo devices in the field and understand how people are using them.
For many companies that offer an Alexa skill, Amazon gets data about the status of a device when you ask Alexa to turn it on or unlock it or whatever. But what few people realize is that Amazon gets state data every time that particular device is used, even if you don’t invoke Alexa. Not every company has to give up this level of data, but many do. Security companies and larger brands tend to be able to carve out exceptions to this level of data sharing.
However, the more than 145,000 different types of Alexa products and the more than 100 million interactions that Alexa has so far handled provide Amazon with a tremendous amount of information about how people live. Amazon first started using this data for Hunches, which it announced in 2018. Hunches allowed Alexa to suggest actions based on the time of day or things happening in the home. When I’d set my alarm before I fell asleep, for example, Alexa would let me know if my August lock was unlatched — assuming, correctly, that if I had turned out my lights and set an alarm that I was likely heading to sleep and so would prefer to have my back door locked.
On Thursday, Amazon announced that developers can enable their devices to let Alexa proactively take action on Hunches. So instead of having a conversation with Alexa about my back door before I head to bed, Alexa would just lock it for me. The user will have to opt into these proactive Hunches and will be able to set some parameters in the Alexa app.
But for developers of smart home devices, this is yet one more tool that takes any intelligence you might have wanted for your device and hands it over to Alexa. It puts Alexa between me and my August door lock, for one, which means that if other door locks work to lock and unlock my door then I might not care which door lock I use because the smart features will come from Alexa. For a smart home device developer whose data has helped Amazon build out such a capability, this might rankle.
Or maybe it allows those manufacturers to focus on their hardware, on building a unique user experience in their own apps (which become much less necessary as Amazon, Apple, and Google take on more of the device onboarding and the smarts). At least that’s what developers tell me. As a consumer, I might really care about home security and thus dig really deep into my lock’s functionality, but I think most of us will pick whatever seems to have the right price tag and work with the right assistant and call it a day.
There are bright spots where products will still be able to offer differentiation, such as in the connected kitchen and cooking products, and likely in lighting, where people want to develop finely tuned lighting designs or schedules. But for many people, the idea of trying to think about your lock, your thermostat, your basic lighting, and security just feels overwhelming and a reason not to buy connected stuff. Having a digital assistant handle much of it for you may make such devices easy enough to work with that the small conveniences they offer become enticing enough to make the purchase.
I also think the standards pushed by the Connected Home Over IP effort will accelerate this trend, while also reducing the complexity that device makers have to deal with when trying to ensure their products work with all of the appropriate smart home assistants.
Amazon also announced two additional services, one of which will cost money and another that is currently free. The first is Guard Plus, which lets additional devices join the existing Ring and other Amazon cameras in monitoring the home and also adds new things that Alexa will listen for, such as a baby crying, footsteps, or motion in the home. Guard Plus takes proactive action if it detects sounds that indicate an intruder, such as triggering the sound of barking dogs on your Echo or turning on your lights. It can also enable users to call emergency services. Guard Plus, which adds to the small number of Alexa-related services, costs $ 4.99 a month.
The second, free service helps people monitor their distant loved ones. Dubbed CarePlus, it lets a user link two Amazon Echo accounts and see when the other connected person interacts with their Echo or smart home devices. Amazon pitched this as a way to stay connected with elderly family members, and it’s not far off from other services designed for this same purpose. Instead of cameras, which people might find intrusive, the services simply let you know that your parent made coffee this morning or opened a bathroom door.
Amazon also lets the user designate an emergency contact, so if the monitored person falls or needs anything they can ask Alexa for help and get connected with the emergency contact. Alexa will notify that person using the Echo device or the Alexa app on the phone. I think these services will be the next field where Amazon hopes to reap valuable data. CareHub combined with health data from the Amazon Halo wearable will give Alexa a good sense of habits and behaviors that might indicate someone isn’t well or needs help.
On the security front, having data from Guard could help Amazon use inputs from cameras, Ring sensors, and more to further refine the algorithms that know when someone is breaking in. Using the physical sensors and cameras plus the auditory data could show Amazon where Guard throws up false positives and so be able to give it a better sense of hearing, as it were. At that point, Guard might become almost the only thing most people need to secure their home.
Even though this event was billed as being about devices, I think it’s pretty clear that Amazon Alexa is really focused on data, and how to use that data to build a layer of intelligence around the home so it becomes ever more in tune with user wants — and at the same time, ever more essential. From there, Amazon can build out new services and even new devices that will help meet people’s needs decades into the future.
The post Amazon’s device day was really about the data appeared first on Stacey on IoT | Internet of Things news and analysis.