Alexa v. Google Assistant makes consumers the big winners

Google was all over The Strip. Photo by Kevin Tofel.

After walking an average of 10 miles a day at this year’s Consumer Electronics Show, I realized I couldn’t walk 25 feet without seeing the same thing over and over. No, not devices; there was wide range of those. I’m talking about products that had Amazon Alexa and Google Assistant integrations.

They. Were. Everywhere. Not just on the show floor either: Google spent a good sum to get its Assistant product mentioned on most hotel signage, the Las Vegas Monorail and billboards everywhere I turned.

That indicates two things to me. First, with such an array of IoT product choices that now have smart voice support, we’re essentially at a tipping point in the smart home market when it comes to mainstream adoption. Second, after years of waiting for de facto, standard smart home platforms that can simplify purchase decisions, we have one. It’s called voice, or what I dubbed the “invisible interface” a few years ago. At a high level voice is becoming both a UI and an IoT platform of sorts.

There are several reasons not everyone who wants a smart home has one yet. Costs have been high and it’s not clear to every homeowner why they even need a smart home, whether it’s one with just a single connected device or dozens. Costs are coming down though and with each new iteration of products, people are starting to see the benefits of having connected door locks, sensors, blinds, thermostats and more.

But the other reason — a main one, I’d argue — is that the smart home market has been confusing for many mainstream people. Ask a non-technical neighbor what Zigbee, Z-Wave, mesh networking or ARTIK is and they’ll probably give a you blank stare. For “normals” to buy into the smart home, all of the back-end technologies and radio protocols need to be abstracted away, never to be seen or talked about again. That’s where voice comes in.

Why? Because if you asked that same neighbor what an Amazon Alexa or Google Home is, they’d very likely know. We don’t know how many Alexa-enabled products Amazon sold in the past two years but consumers did purchase “tens of millions of Alexa-enabled devices” this past holiday season. Likewise, Google sold an estimated six million Google Home Minis in the final three months of 2017. People are buying these because they provide instant benefits and are intuitively simple to use simply by asking questions. And they are buying them: NPR says that 16 percent of U.S. households now have a smart speaker, which is a 128 percent increase from NPR’s data a year ago.

Many of the newest products I saw don’t require hubs either because they’re working natively with a voice assistant. Going forward, you’ll see Samsung’s Bixby voice agent in televisions and refrigerators. Televisions from LG, Sony and others can be voice controlled directly through Alexa or Google Assistant without a hub. In fact, Google announced this week that its Assistant / Home platform works with more than 1,500 devices.

So it’s becoming less important to know which smart home products work with Wink, SmartThings and other branded-hubs because voice controls are essentially becoming the newest and primary interface for smart home products. Sure, for many things you’ll still need a hub. If you want Bixby on to show who’s at your front door from the fridge or TV, your video doorbell will have to work with SmartThings. The same goes for a notification that you left the garage door open or that your home security system indicates a family member just arrived home. That isn’t going away.

But that’s OK. By integrating now-standard voice platforms into a larger array of smart home products, consumers will have an easier time understanding the value and in installing or using connected gear. No longer do we have to worry about news like Honeywell’s announced integration with Whirlpool that lets your Honeywell thermostat tell your appliances when you are out of the home so the dishwasher can run. We can just hook each to the Amazon Echo and tell it to turn on the dishwasher as we leave. Automations may require a hub or a third-party service such as IFTTT, Yonomi or Stringify but getting smart devices working by voice in the home is an important and simple first step.

I definitely don’t want the “Works with Wink / SmartThings / Nest” designations on smart home products to disappear. However, the “Works with Amazon Alexa / Google Assistant” markings have become far more important for mainstream consumers. Let the two (or three, if you include Bixby) voice assistants continue to battle it out, I say. In the end, we all win.

Stacey on IoT | Internet of Things news and analysis

ARM’s CEO on Spectre and Meltdown, plus hot CES 2018 takes

Google was really pushing Google Assistant and the Google Home.

This week. the Internet of Things Podcast crew went to CES to discover that the consumer electronics industry was ALL OVER the internet of things. Yes,  Amazon Alexa and Google Assistant starred in everything, and we explain what that means.  Plus, we answer a question about bathroom fans taken from the listener hotline. After all that Simon Segars, the CEO of ARM, kindly talks about how to be safe with the massive Spectre and Meltdown security vulnerabilities and gave some context about what this means for the internet of things. Listen up.

Stacey on IoT | Internet of Things news and analysis

Leak detection: 4 startups worried about water

The Buoy system tracks your water use and provides a remote valve shut off function for $ 799.

Leak detection sensors have been around for years, but most have limited sensing capability. If the sensor isn’t where the puddle is, the leak goes undetected. Some leak sensors have tried to address this with corded “tails” that extend beyond the sensor but can also detect water. I like those better, but the best leak detection isn’t a smattering of sensors. Instead, it’s a smarter water meter.

With that in mind, several companies have formed to try to help consumers and insurers avoid one of the costliest aspects of home ownership — undetected leaks. At CES next week, several startups will launch or show off new products aimed at the DIY homeowner that are worth a look.

All of them allow for remote shut off of the water valve if a leak is detected. Some even do that automatically. Most offer algorithmic leak detection, which means that water doesn’t have to hit a sensor for a leak to be detected. Instead these systems monitor the water usage of the home looking for anomalies or above average water usage by specific devices. Toilets are a big culprit.

Toilets that run all the time can waste hundreds of gallons of water each day in a home, which adds up over time. For example Phyn, one of the startups in this space, determined during a beta test that just over half of leaks are caused by toilets that run constantly. Other common culprits are sprinkler heads and faucets.

Phyn is probably the most well known of these startups to this newsletter’s readers because I covered the creation of the company in 2016. Phyn is a joint venture between Belkin and Uponor. Uponor makes plumbing and heating supplies. Phyn’s product uses Wi-Fi and sensors that determine how water flows throughout the home and when those flows deviate from the norm. It has been in beta in a few hundred homes for over a year, and will sell for $ 849 plus installation.

Flo is another startup in this space, offering a $ 500 device that will ship in February. The Flo device screws onto the home’s main water line and tracks the water flowing through the pipes. It also sends out proactive tests to try to detect slow leaks, something that Phyn says it can also detect.

The Guardian by Elexa is a bit different because for $ 400 you get a motor that screws onto your water shut off valve to turn it if a problem is detected. Also included are sensors that you scatter around the home to monitor for overflows or moisture. This is not my favorite approach since sensor placement will determine if your leak gets detected or not. Of course, even the best algorithmic leak detection sensors won’t help in the case of roof leaks or some of the other ways water can destroy a home.

Finally, there’s Buoy, which is $ 800. Buoy also installs on the main water line and uses algorithms to determine what’s using water in your home and what’s normal. On its website, the company also gives the example of a running toilet wasting water.

All of these startups are hoping consumer panic about leaks helps drive sales, but the real test will be insurance providers. Already insurers are testing these devices and we’ll likely see discounts if a consumer installs one of these in her home, or even certain policies that might cut premiums but force such a device on the user. If it can detect a slow leak, I’d be all in. Those are expensive!

Stacey on IoT | Internet of Things news and analysis

Internet of Things News of the Week, January 8 2017

Mapping the floor, and your WiFi

Want this IoT news summary in your inbox every Friday? Subscribe to the newsletter.

What exactly is going on with all of our chips? Security researchers have discovered that a flaw in Intel and some ARM processors can leak protected data if exploited. There are two exploits, Spectre and Meltdown. What’s worse is that the solution to Spectre isn’t readily available. OS vendors have tweaked their operating systems to mitigate the risks associated with Meltdown. As far as security news, this one is huge and people are trying desperately to figure out what is going on and how to fix it. For the more technically minded, the tweet stream is a good place to start. If you are less inclined to dig into how computers handle memory buffers then read Ars. (@gsuberlandArs Technica)

Layoffs at Eero: Mesh Wi-Fi startup Eero laid off about 30 people — or about 20% of its workforce — according to this article. Eero confirmed laying off 30 people, and said it was killing a speculative project so it could focus on its “core business.” My hunch is that Eero realized that times are getting tougher for once high-flying IoT startups and it should focus on getting revenue and building a business. Especially since its claim to fame (mesh Wi-Fi) has been copied by other companies. (TechCrunch)

Speaking of Wi-Fi: Roomba vacuum robots that have Wi-Fi chips will soon use their radios to create a map of Wi-Fi coverage in your home. This way you can avoid dust bunnies and dead spots. After Roomba’s CEO got in trouble a few months back for proposing that the company would map users’ homes, this features sounds more like an attempt to do something … anything … with some of the capabilities on the device that don’t involve selling user data. After the hubbub over mapping, the CEO promised never to do that. I’m curious who takes them up on this.  (TechCrunch)

Yes, even more Wi-Fi news: Cirrent, a startup I’ve been excited about for quite some time, has signed on several big name brands to use its automatic provisioning service. Electrolux, Cypress and Ayla Networks are now using Cirrent’s ZipKey service to get devices on the network faster. Electrolux will use Cirrent for its appliances, Cypress will make the feature available on its silicon and Ayla will let any company using its cloud offer the functionality. ZipKey works by using an available and certified Wi-Fi network to make a connection with a product when it comes out of the box. Then a consumer can claim the product and move it over to their own Wi-Fi network. Rob Conant, CEO of Cirrent, says ZipKey now has Wi-Fi networks covering more than 120 million homes in the U.S. and Europe — or 40% of them. Comcast is one of the companies providing hotspot access for ZipKey.

10 IoT companies to watch: The close of the year is a perfect time for lists and predictions, and most are pretty redundant. However, I liked this one from EE Times that starts off with nothing great, but then redeems itself by digging into the idea that many middlemen in the IT ecosystem are trying to absorb more roles in the industrial IoT. As an example, Arrow (a distributor and owner of EE Times) buying a company this week that lets it take on systems integration. The startups are pretty good too. (EE Times)

Smart baking startup gets more dough: Sorry, I couldn’t resist. Drop, the maker of a connected scale, raised more than $ 7 million in VC funding from firms like Alsop Louie Partners, according to an SEC filing. Drop has parlayed expertise from building its connected scale into creating recipes designed for humans and machines to make things together. That sounds fancier than it is today, but through a partnership with GE, you can use a Drop recipe to preheat your oven at the right point in a recipe automatically. Other automations could follow as cooktops get smarter.  (Axios)

Ads on Alexa? Eek! This week CNBC reported that Amazon was discussing marketing opportunities with large consumer product companies. My colleague Kevin wasn’t thrilled and started wondering what would make a good voice ad on a smart speaker that could be used by multiple people. Hint: Not much. (StaceyonIoT)

New York may be the first to try to hold algorithms accountable: I starting thinking about algorithmic bias in 2015 after I left Gigaom. I toyed with the idea of doing a fellowship or writing a book about how programmers were building a world that was data-driven and seemingly rational, but was fueled by all kinds of assumptions in the code. That world is rapidly coming to pass, and even without a book, people are wising up to this. As more sensors track more things, we’re going to get a lot of junk algorithms trying to nudge people in particular directions. I loved this story of a New York City Council member trying to ensure those algorithms are transparent to citizens. (The New Yorker)

A second neural network on a stick: Last year I got excited about Intel putting silicon from Movidius on a USB stick. The idea was to offer a powerful computer vision processor in a mobile format to see what people would do with it. To me it represented the beginning of being able to train neural networks at the edge. Now, there’s a second neural network on a stick from a startup called Gyrfalcon Technology that claims to be much more powerful and use less energy. The future is coming faster than I thought. (Alisdair Allan)

Stacey on IoT | Internet of Things news and analysis

The trouble with the industrial IoT is too much IT

DOUGs keep the lights on.

Much has been made of the difference between information technology (IT) and operational technology (OT) when it comes to the industrial internet of things. The big tech companies have come into the space with their gateways, clouds and wireless networks, sometimes without understanding that for the OT world, automation has been happening for a long time.

In many cases what’s new isn’t the sensors, but the data analytics and ability to start reacting in real time to that data to add more automation. As the IT world grasps what the industrial world wants, providers are spending more time discussing edge computing, security and service level agreements.

But when I travel between the two worlds, people are still baffled by the industrial guys. These are the operational technicians who keep parts functioning, lay wire and generally ensure that all the bits and bobs of whatever automated system continue to work. These technicians have been doing this for decades. It’s what they are paid to do: keep the plant operational.

They frustrate the heck out of any new manager hoping to come in and add mobile apps or IoT analytics to the floor. In a conversation with one such executive who manages a team of OT and IT staffers, he derided the operational guys as DOUGs. It stands for dumb, old utility guys. As in, we wanted to see if we could get our sensors to report an additional data point, but the DOUGs didn’t want us to touch them.

I laughed, but I realized that it wasn’t the DOUGs that were the problem. Litterally their entire job is to build something that works and keep it running no matter what the world throws at it. If you are a lineman for the phone company or a utility and a line gets knocked out, you grab a truck and put it back.

On a manufacturing floor, you add new devices and processes slowly with an eye toward ensuring that nothing breaks and the end product isn’t screwed up. Intellectually we get this. But that is why DOUGs are such a pain in my friend’s butt. Operational guys keep things operating. They don’t think up new features or strive for continual deployment of new services.

They keep the lights on.

Meanwhile in the software world, those who “move fast and break things,” get ahead. Their focus is on agile delivery of new features. Even the emphasis on the cloud a little over a decade ago was all about freeing up operational resources so the IT staff could move quickly and adapt.

Think about your favorite software. I bet it has crashed at some point. Or your last Skype call. Or the last time you had to reboot a modem to get things working again. It’s not to say that technology doesn’t work, but when things fail, the goal is to fix it and get back up again quickly without too much fuss. And the repercussions for failure have historically been low.

But as IT meets the real world, the repercussions for failure are much higher. Consider the real-world effect that computer issues can have on the airline industry — halting travel. Or consider Chrysler’s software error that could prevent airbags from deploying and led to a massive recall in 2017. For as much as the IT world values continuous improvement code, disruption and agility, when the code hits the road lives or livelihoods are at stake.

So as frustrating as the DOUGs are, as we push more IT into OT processes, it’s worth viewing their perspective. Sometimes changing something isn’t as useful as just keeping something up and running. Sometimes moving fast and breaking something is a really dumb thing to do.

And for those in charge of trying to blend IT and OT, maybe understanding the DOUGs will help you succeed in your next project.

Stacey on IoT | Internet of Things news and analysis