In a connected world, hardware is a Trojan Horse

Facebook this week said it had hired Regina Dugan away from Google. Dugan led the Advanced Technologies and Projects lab at Google, which tried to build the impossible. Some projects included a modular cell phone and sensor-infused fabric. Facebook isn't a stranger to hardware, having built and designed its own servers, but it's significantly raising the stakes.

With Dugan's hire, it is creating a hardware effort called Building 8 that seeks to build consumer hardware. Facebook spokesman Michael Kirkland explained that Facebook sees a lot of gaps in bringing technology such as virtual reality to the mass market. Other areas of interest involve gear for better and faster internet access. Kirkland specifically mentioned Project Soli, which uses radar to “see” small gestures and turns those gestures into a signal a computer can understand. Soli is designed for capturing movements as small as a finger flick. Such fine-motion capture could make the VR experience more like real life.

"There's a lot more we're going to try to do to to fill in some of the gaps in the ecosystems we're trying to create," says Kirkland.

At F8, Facebook's developer conference CEO Mark Zuckerberg offered a 10-year roadmap that requires advancements in artificial intelligence, internet connectivity worldwide and virtual reality. "There's a huge punch list that needs to happen in hardware and software to get to that end-state," Kirkland said.

So what does this have to do with the internet of things? Facebook is attempting to use its enormous resources to change how we communicate. Facebook's goal is to connect people worldwide. But to do that it needs to build technology and tools that don't exist yet. But it's clear that Facebook isn't viewing its hardware as its primary value. It's just a means to an end.

And that mindset is crucial for entrepreneurs and companies investing in the internet of things. A lot of people think the hardware they are selling is their primary value. But in a connected world full of cheap sensors, the hardware a business sells is more often, just a way to solve a problem. Solving the problem is the real value a business provides. Understanding that is essential to recognizing the current opportunity facing a business, and adapting as the market changes with the internet of things.

For example, Inder Singh, the CEO of Kinsa, a connected thermometer company, saw that there wasn't reliable information on where diseases were and how they might spread. Governments and health officials relied on patients coming into hospitals or doctor's offices to report their symptoms, and then on the doctors to report those patients to an organization like the Center for Disease Control. Singh recognized that the problem he was trying to solve was real-time access to high-quality data about the spread of illness.

But to solve that problem, he needed hardware. So he built a connected thermometer that collects data on the spread of fever, which is a reliable sign of illness. To back that up, there's an app that lets people report symptoms as well. Sales of the thermometer are doing well, but the real value Singh has to offer is data. After reaching a critical mass of thermometer customers in Austin, Texas, Kinsa launched a "weather report" on various illness.

A lot of companies talk about digital transformations, but recognizing that the actual problem you are solving is probably closer to your real value, is essential for success in the coming years. Hardware is going to become rapidly commodified. As that happens, the sources of data will also become commodified.

The promise of the internet of things isn't just that you can sell a more expensive widget that can be controlled by a smartphone, it's that the access to real-time data at a potentially huge scale can provide insights that solve pressing problems. Problems we couldn't solve before. Recognizing that those insights are the real value your business will offer, and not the device itself, is a cognitive leap that a successful company has to make. Otherwise, it can lose focus.

And rest assured a competitor will be right behind it ready to take up its slack.

Startup profile: Deako's bright idea

Deako is a Seattle-based company that has just raised $1.8 million, bringing its total to $3.3 million in venture capital. It sells builders a connected lighting system that consists of a hot swappable light switch that fits into a standard electrical box. The switches can offer homeowners everything from fancy, gesture-based controls to a simple dimmer, depending on which switch is used.

You could install a fancier version of the switch by your door that not only controls your porch lights, but also has a spot you can press to turn out all of your lights. In other rooms, you might opt simply for a dimmer.

The idea is for builders to install these switches during construction, ideally in planned communities. Builders can sell the fancier lighting controls as a premium upgrade, much like people upgrade to granite countertops. The system resides in the home, and can also be changed over time as the technology inside the switch becomes outdated.

Making the switches easy to replace without worrying about changing the electrical wiring means that if the tech inside fails or becomes obsolete, the fix is an easy replacement.

Derek Richardson, Deako's CEO and cofounder, doesn't currently have plans to go beyond lighting, or to connect the lighting to other systems installed in the home. He thinks simplicity is the way to go to get mainstream adoption and builder buy-in. In many ways, this is a smaller step to get to a smart home. There are no algorithms or a hodge-podge of sensors. But because Deako is targeting builders and thinking about issues such as longevity, it may be taking the right approach.

Podcast Episode 54: Fashion meets the internet of things

Do you want your smart jewelry to have a screen? If we have multiple pieces of connected jewelry how do you make it easy to program for the day? Or will you only have one sensor-laden wearable akin to a smart watch that does everything? These are some of the questions Matt Manley of Fjord tackles with me on this week’s show. We start off discussing jewelry, but veer off into how devices should deliver ambient information and the state of wireless power. Even if you aren’t into wearables, Manley’s comments on notifications are worth a listen.

Kevin and I kick off the show with jewelry as well, discussing the newly launched Aries bracelet from Ringly. We then talk about the $12.5 million in funding for Luma, one of the companies trying to make a mesh router. This one offers parental controls and should be out in April. We also take a look at the Wirecutter’s review of the best smart switch (outlet). For those of you shopping, they liked the Belkin Wemo Insight Switch. We quickly discuss Pfizer’s plan to use existing sensors to monitor Parkinson’s patients and the lifesaving Fitbit data everyone was so excited about. And like the rest of you on SmartThings, we’re waiting for a fix of the system which has been broken for almost four weeks.

Listen to this week's show.

Support.com, which provides customer support for at least one connected home platform, issued a report in March that had some worthwhile data about where people buy connected products and what they expect from them. There's clearly a gap between what people want, and what people get today with connected products. (Support.com)

IoT news of the week

Hitachi will spend almost $1 billion on IoT: Hitachi, the Japanese conglomerate, is setting up an internet of things research center in Santa Clara, California. The center will open as early as the first half of May, and Hitachi plans to invest roughly 100 billion yen ($915 million) in IoT over three years. Hitachi President Toshiaki Higashihara said the company had to make this investment so it can build new business models. It must do so because it needs to boost its operating profit margins to 10% or higher to compete with rival GE. (Nikkei)

Want to be an IoT professional?: I stumbled across a new member-based organization dedicated to training the next generation of IoT professionals. Cisco is heavily involved, and the goal seems to be some kind of certification and training effort for the internet of things. Certifications can be helpful, but they can also be a money grab for the organization that handles the certification process. I'm not sure we have enough standards or even a technology stack that makes an IoT certification valuable just yet, but there is a nice listing of available IoT-related jobs at large companies on the site under the talent exchange section. (The IoT Talent Consortium)

SmartThings offers a fix for frustrated users: On Thursday, SmartThings let users know it had tried to fix the troubles its customers were having running routines and its Smart Home Security Monitoring service. In an email to users, it said it has moved everyone to a new backend scheduling database that runs routines and scheduling and it "tried to enhance" that database so issues don't happen again. Some users report improvements, but not all. For a deeper dive on the infrastructure woes, which stemmed partly from the proliferation of third-party apps being used on the service (those apps are usually a feature, not a bug) click through to the link. (SmartThings Community Forum)

Philips Hue gets a facelift: On April 20, Philips is updating its app to add a security certification for third-party apps. The good news is most third-party apps already have worked with Hue to ensure their apps work post-update. The bad news is there is a small list of apps that will stop working if users do a firmware update or clear their cache on their Android device. Hue has provided a list of these apps on its FAQ site. For fans of other light bulbs that work with the Hue system, those should still work if they are ZigBee Light Link compliant. Users will also notice that the firmware update changes their existing Hue app icon. This is because soon Philips is launching a new app, and wants people to be able to differentiate between them. (Philips)

A good idea for smart home data: We're creating a trove of data as we add connected devices and sensors to the home. A UK designer has considered what data should belong to you and what data should belong to the home. She suggests that the city should keep this data, but the framework is interesting for insurers, homeowners and Realtors. It's actually close to efforts made by the technology lab at the National Association of Realtors, which would like to see this data put into MLS listings. (The Good Home)

Will the internet of things kill network neutrality? Telecom expert Martin Geddes has been telling regulators and anyone else who will listen that "best effort" telecommunications networks won't cut it. Now he's using the internet of things, and the services that connected devices and ubiquitous data can enable, as a new way to argue that current network ideas will become obsolete. I'm not sure I agree with him, but his description of the promise of the internet of things is right on target, and it's worth recognizing that the reliance on the network for the internet of things means debates over how data will flow over those networks. (Martin Geddes)

A sensor will soon replace a human's job in Massachusetts: And actually, that will be a good thing. A new law means fishermen in New Bedford, Mass. will soon have to bring a human monitor to ensure they don't overfish. The person must be paid and must go on every trip they make. This will hurt smaller fisherman, so the city is hoping that Massachusetts company Inex Advisors can build a sensor that could quantify and identify the catch. (Fortune)

UL wants to certify connected devices as secure: But some outside security researchers think Underwriters Lab isn't being transparent because it wants to charge researchers $800 for a copy of the certification plans. There is also a question of whether or not you can certify an ever-updating device and whether a pass/fail certification is the right way to go for consumer connected devices. Others favor a nutrition-style label that offers granular detail about the potential risks. IMHO this style of label is also how we should be communicating with consumers about how their data is used. (Ars Technica, UL)

Connected devices give patients the power: There's an article in JAMA all about The Nightscout Project, which was an effort by parents to hack their diabetic childrens' proprietary medical devices. I encountered The Nightscout Project at South by Southwest in 2015 and was impressed by the parents' efforts to modernize products that completely ignored the advent of smartphones and the cloud. The JAMA article brings up another element to this, that connected devices, patient activism and some code can also create a bottoms-up medical product that serves users better than anything doctors or the medical community could design. (JAMA, The Nightscout Project)

Techstars has a new IoT accelerator: Techstars has teamed up with Verizon, SAP, Bosch, GE and PwC to create a New York City accelerator for the internet of things. Ten IoT startups will be selected to participate in a 15-week program. The program will focus on pretty much everything this newsletter focuses on, plus virtual reality. Mentors include folks from PwC, GE, SAP, and Bosch, as well as people from startups such as Filament (a blockchain-based sensor network) and Sphero. Design guru John Maeda from Kleiner Perkins is also a mentor. (Techstars)

So what makes a good accelerator?: The number of U.S. accelerators has increased at an average rate of 50% each year from 2008 to 2014, which is apparent to anyone watching the startup sector. What's less apparent is how good those accelerators are. The Kauffman Foundation is trying to rectify the lack of quantitative data with studies but also did a round up of existing research on accelerator effectiveness. So if after your Kickstarter you plan to submit to Y Combinator, TechStars, Highway One or another program, check this research out. (Kauffman Foundation)