How to fix the fragmented smart home without standards

There may never be one app to rule them all. But maybe there will be a voice-activated assistant?

Last week I went to San Diego to visit Qualcomm and to attend CEDIA, a trade show for professional AV and home automation installers. The narrative flowing through both events was that the internet of things as it relates to the smart home is a jumbled mess. (It’s a jumbled mess elsewhere, but that wasn’t the focus of last week.)

Unfortunately the response to this mess hasn’t been a turn toward standards or soul-searching among the major technology companies, who are seeking to divide things further with incompatible systems such as HomeKit or proprietary radios. Instead, the response as seen from CEDIA was this idea of bringing in installers and experts to pull together a smart home for users and then let homeowners control it from their smart phone.

People call this the Do-It-For-Me model. Companies have adapted to it with things like Best Buy’s Geek Squad, Amazon’s launch of expert installers and even the introduction of products like the Ring doorbells and August locks for the professional installer channel. The smart home market is essentially shrugging its shoulders and saying, “Yes, this is hard, so just pay some money and someone will handle the rough bits.”

The business models will still have to change for custom installers, and consumers will likely find themselves paying more, but these changes will result in the gradual move of home automation down market.

Up the road from the San Diego Convention Center, Qualcomm is solving similar issues with two pieces of technology. The first is a highly integrated system on a chip (SOC) that combines three radios on one board called the QCA4020. The QCA4020 was launched in February, and the promise is that if someone puts this into their connected product then the user doesn’t have to worry if their stuff has Wi-Fi, ZigBee/Thread or Bluetooth.  The device will connect to anything using those radios.

This is already happening in some cases, thanks to the Amazon Echo. It acts as an aggregator of radios and controls devices that work in a variety of systems, but even voice interfaces are frustratingly bifurcated. This is where Qualcomm shows off something so compelling that I want it now.

The company showed a demo of its audio processing technology using the Amazon Echo, Google Home and Apple’s Siri. What was compelling is you could ask one named entity, such as Alexa, a question but then determine which voice assistant actually responded. (This happened in the demo by manually selecting which platform would respond.) So I might say “Alexa, can you put mangos in the fridge?” while selecting Google to actually answer the question.

I’d do this because Amazon’s Alexa isn’t very good at answering these types of questions, but Google assistant is. Unless you have two assistants side-by-side in your home, getting the best response can be hard. Qualcomm’s software appears to mitigate this. I’d love to see the ability to assign specific tasks to Google, Alexa or Siri in an app so I don’t have to do it every time I ask a question.

Amazon clearly sees a need to broaden its capabilities, hence its agreement to integrate with Microsoft’s Cortana. However, it’s not clear if Apple or Google will happily join this cause. That’s where efforts like Qualcomm’s come in. However, the timing for any such help and the form factor remain to be seen. Qualcomm didn’t have an anticipated date for this voice assistant bridging software.

Sadly, we’re at the point in a technology cycle where the tech isn’t holding us back, so much as the power and jockeying of the major tech companies that don’t want things to work together. To bridge disparate radio standards, devices and even assistants will require a commitment among tech players to create a flexible user interface built around an AI that can talk to everything.

Historically, you could get away with a commitment to one computing platform (Mac or PC, iOS or Android) but when we’re embedding tech into everything that model makes less sense. Your choices in home appliances, light bulbs, vehicles and watches shouldn’t be limited by the phone you carry or the personal assistant in your smart speaker. Those will all change over the life of some of these larger products.

Ideally, the same benefits our smartphone apps have gained through cloud APIs could be applied to our assistants. Everything we want connected is connected, and we can use the invisible interface of voice with any service we want, regardless of device choice.

Stacey on IoT | Internet of Things news and analysis

Nest debuts Nest Secure home system and has a Thread router called Nest Connect

It’s almost hard to believe that then Google, now Alphabet, bought Nest in January, 2014 when looking at Nest’s evolving product line since then. There’s been very little notable Google influence. With several new hardware products announced today, however, that’s changing.

At a press event on Wednesday, Nest introduced another outdoor camera as well as a security system and doorbell, and announced the addition of Google Assistant features to a current camera, while further supporting both the Weave and Thread protocols.

Interestingly the Thread Group noted something that Nest didn’t say much about: A Thread-based router called the Nest Connect. Here’s a picture of it from the Nest press site.

In an emailed statement, Grant Erickson, president of the Thread Group said, “Nest Detect relies on Thread’s low power architecture to last for long periods of time on one battery. Nest Connect is a Thread router, which seamlessly extends both the reach and connectivity of the Thread network. Nest Guard provides seamless IP-based routing between Thread, Wi-Fi, and cellular, and their interoperability with the Yale Linus Lock is the posterchild of Thread’s benefits in action.”

That sounds like a potential method to address the security challenges of using a traditional Wi-Fi network or router to secure IoT devices but we’ll have to dig for more information since Nest didn’t mention this.

Let’s move on to what Nest did announce.

First up for owners of the currently available Nest Cam IQ security camera that launched in May of this year: With a software update coming “this winter”, you’ll be able to speak to Google Assistant through the camera. No, you won’t be streaming music through the little speaker in the camera but you can ask for information or speak commands to your other connected devices such as lights and doors.

New to Nest’s lineup is an outdoor Nest Cam IQ Outdoor which is IP66 rated for dust and moisture. This unit isn’t battery powered so you’ll need to hardwire it through the walls of your home. Like its indoor namesake, the outdoor camera has IQ technology to recognize people and alert you when it recognizes someone. Pre-orders start today for $ 359 and the device will begin shipping in November in the U.S., Canada, and Europe.

The Nest Hello doorbell is also new but won’t be arriving until early 2018 and no price was announced.

As you’d expect, it has a video camera (with High Dynamic Range support!) as well as night vision, a 160-degree field of view and a speaker so you can remotely see who’s at the door and speak to them from the comfort of nearly anywhere. The Nest Aware subscription service can notify you when a particular person is seen by the doorbell.

Lastly is the Nest Secure alarm system, which is comprised of several components and available in November in a starter pack for $ 499.

Nest Guard is the Wi-Fi-enabled heart and hub of the system that also acts as a keypad for arming or disarming, a motion sensor and “a friendly voice” as well as an 85dB siren. You can monitor the system yourself or pay a monthly fee to MONI for them to monitor it, just like a traditional security company would do.

Ideally, you’d place Nest Guard near the exit to your home similar to any centralized alarm panel. Included in the starter pack are a pair of magnetic Nest Detect window or door sensors and two Nest Tags: NFC Fobs to disable the system without tapping a passcode. Nest will also be offering a cellular backup service for Nest Secure in the future for either $ 5 monthly or $ 50 per year.

The system is smart in a similar fashion to the original Nest thermostat. It can ping you if it realizes you left home and didn’t arm the security features, for example. You can also leave the security system armed but open a door — say to let a pet outside — without tripping the siren: There’s a button on the Nest Detect sensors for this situation. And walking past a Detect at night fires up a small nightlight on the bottom of the sensor so you don’t trip on Fido’s toys.

All in all, the new product line is what I’d expect from a smart device company: Adding intelligence and convenience to traditionally “dumb”, mundane systems. Nest says its new products will seamlessly work with the Nest mobile app. It also touted the Works With Nest program as a way for other device makers supporting Weave and Thread to integrate with the system in the future.








Stacey on IoT | Internet of Things news and analysis

SecureRF is bringing crypto to the embedded world

Louis Parks, the CEO of SecureRF.

SecureRF makes software that gets embedded onto microprocessors that allows them to perform impressive feats of security in milliseconds, not minutes.

Since security is one of the current challenges faced by the internet of things, a solution that can handle the complex computing tasks required for public key encryption without requiring a powerful processor is exciting. How exciting? I first met the company last year at ARM Tech Con and thought it was pretty interesting. Then, two months later I was discussing IoT security over dinner with the CTO of a large industrial and defense company who raved about what SecureRF was able to do.

The CEO of the company, Louis Parks, explained the details to me over the phone. Unfortunately, a lot of his explanation involves the complicated kind of math people start throwing around when they discuss cryptography.

The net result of this math is that SecureRF’s crypto works well even on 8-bit microprocessors and can perform the calculations needed to authenticate a remote device in a fraction of a second. And while it’s possible to run other cryptographic schemes on a tiny microprocessor, it would certainly take longer (think minutes, not milliseconds), making it bad for handling things quickly.

The real question is how often companies need to do public key encryption on relatively dumb devices. Going forward, I bet it will happen a lot. Payments from wearables are one example. There are also enterprise and industrial use cases.

A few episodes ago in the Internet of Things Podcast, I spoke with Alisdair Allan, an academic, who is proposing a new model for securing the internet of things.

His focus is on efforts to break a system by inserting false data. His example was a faked soil moisture sensor that cost a vineyard hundreds of thousands of dollars. A hacker created a fake sensor that told the irrigation system the soil around it was dry. Then, the irrigation system, hearing the problem, turned on the sprinklers to solve it.

Since there wasn’t a need to water, the false information led to fines for using too much water during a drought and a poorer crop. Allan’s proposal is to use blockchain technology to establish histories for the reliability of every sensor in an ecosystem and then use that information to help an actuator make a decision: Basically to teach machines that sensors may lie, and use the blockchain to give them a way to establish trust.

Putting a crypto key on a sensor to establish that trust is complicated. However, SecureRF is working with chip companies to get its algorithms and software embedded on established microprocessors. It also works with customers to design specialty chips that will run the algorithms as part of an overall solution.

As both security and IoT continue to gain attention, SecureRF has technology worth looking at.

Stacey on IoT | Internet of Things news and analysis

Why do enterprises want IoT?

Connected cars, workers and buildings offer almost complete visibility into business operations. But for the most successful companies, the process starts with a use case. One executive at a prominent company doing IoT deployment told me, “When someone calls asking for ‘some IoT’ I don’t even call them back.” I imagine he does call them back, but he’s right that he can’t help until he has a use case. So what are the big use cases in enterprise IoT deployments?

Verizon this week issued a report on its view of the internet of things featuring drones, use cases and more good information. But my favorite was the chart below, discussing what executives were most interested in.

Stacey on IoT | Internet of Things news and analysis

Chamberlain plans to charge for IFTTT, Google Home integrations

Some of the possible options for the MyQ and IFTTT integration if users want to pay the fee.

Chamberlain, the maker of the MyQ garage door opener and enterprise access products, will this week start charging for select smart home integrations, including IFTTT.  Chamberlain will charge customers $ 1 per month, or $ 10 per year, so they can link their garage door openers to their other home automation devices.

Cory Sorice, VP and general manger of emerging business at Chamberlain, said via email that the new service will launch on Tuesday and will cover IFTTT and Google Home integrations. He also said the company plans to roll out more partners over time. I believe Chamberlain is the first company to try to charge an additional fee for linking a connected product to a service like Google Home or IFTTT.

When asked why Chamberlain was charging a fee for integration, Sorice said that the market has evolved into so many options for controlling the smart home, including HomeKit and Google Home. These require additional support. Additionally he said users are fragmented between those who want to embrace higher levels of home automation through outside services, and those that just like using the MyQ to ensure their garage door is shut after they leave.

As someone who has purchased the MyQ and long been frustrated that it didn’t work with the Amazon Echo and Google Home (it does work with Wink), this news is bittersweet. As a consumer, I don’t want to pay more to link my garage door to other devices, but as a reporter covering the industry I am well aware that for a connected product to remain sustainable, companies have to charge a lot up front or come up with additional services to pay for the ongoing cost of software development, cloud computing and security updates.

Cloud storage for video cameras and doorbells or monitoring have been the most common way connected home companies have tried to keep revenue flowing in, although there are other options. For example, your thermostat company might sign a partnership with an air filter replacement vendor to remind you to buy a new filter and let you order it from within the app. At that point the thermostat company could take a cut of the air filter price.

Chamberlain’s plan is another way to add a recurring revenue stream to support a product that will carry development costs for the rest of its life. One of those charges is actually the IFTTT service itself. Earlier this year IFTTT rolled out two partnership programs that companies buy into. One is $ 199 a month and the other starts at $ 499 a month depending on the size of the company. Given Chamberlain’s focus on security and delivering a high quality service, it’s likely in the pricier tier that offers guaranteed servers and tech support.

I emailed IFTTT to discover if other companies were planning on charging for integrations, and a spokeswoman got back to me saying: “It’s an interesting move and our team is watching closely. … It’s exciting to see Chamberlain take ownership of their IFTTT integration and work with their customers on the value of MyQ Applets.”

Devices like Philips Hue light bulbs, SmartThings’ home hub and August locks are popular home automation integrations. Could  the companies behind those services start charging too?

As for Chamberlain, ahead of the official launch of the $ 1 service fee, Sorice said while some users are clearly frustrated on Twitter, the feedback has been generally positive and uptake has been ahead of expectations. For most consumers, an extra fee means they won’t link their service to IFTTT without having a clear use case. And frankly, given the difficulties that people have had finding a compelling use for the smart home, and the costs of supporting connected devices over the long run, this fee might be a positive step for the industry.

We discussed this topic and more last week on the Internet of Things Podcast around the 13:20 mark, if you want to hear more.

Updated: This post was updated at 2:30 CT to add comment from IFTTT. 



Stacey on IoT | Internet of Things news and analysis