Some folks are scratching their heads over Google’s intent to spend $ 50M to purchase Xively from LogMeIn. I’m not one of those folks because Xively quickly gives Google a few things that it’s lacking in this market.
It was just a few podcast episodes ago where Stacey and I were wondering, “What and where is Google’s IoT product strategy?” At the time, we didn’t see a cohesive message or product toolbox from Google like you can find for Microsoft, Amazon or IBM. To be fair, Google does have its Cloud IoT Core, which has similar capabilities to Xively, which provides an end-to-end IoT solution including device management, application support, service integration and data analytics.
Sure, Google’s Cloud products can be mashed together for all of that as well. And Google is excellent in the areas of software integrations and analytics. Device management and deployment though? There just isn’t enough history here for Google to justify saying yes. Keep in mind that the Google Cloud IoT Core was announced in May 2017 at Google I/O, so it’s not even a year old yet. Maybe that’s why when I tried to find some customer stories and case studies for it, I came up empty.
Compare that to Xively, which is ancient by IoT standards: The company launched as Pachube in 2007 and in 2011, its platform was used to connect geiger counters across Japan after the Fukushima Daiichi nuclear disaster caused by an earthquake. That same year LogMeIn purchased the company for $ 15M, rebranded it as Xively and began to expand the customer base.
In fact, on Xively’s site you’ll see customer case studies that I would have hoped to have found on Google’s Cloud IoT site: Customers such as Lutron, ShadeCraft Robotics, and Heatworks, to name a few. Lutron’s story is particularly interesting since after deciding to build connectivity in its products, it took just four months “from concept to field-ready product.”
That’s equally as important as the platforms and services Xively has created over the last decade because while Google knows its own products better than anyone, it doesn’t always have the end-user customer experience to understand how its products are used. Yes, Google is great about asking for feedback. But working with industrial and commercial IoT product makers requires a more personal touch: Xively has a Professional Services group providing insight from the beginning to the end of an IoT project.
Obviously, there’s no guarantees that a Xively purchase will give Google the IoT boost it needs to better compete against Microsoft, Amazon and IBM.
But if you nose around Xively’s site like I did and then compare it to Google’s own IoT product sites, you’ll see that this is a big step in the right direction. Not only is there a cohesive messaging strategy but there are platforms, services, products and experienced people to help Google deliver on its IoT dreams.
Get out the guacamole, because you’re going to hear a lot about chips on this week’s Internet of Things Podcast! ARM announced a new architecture for machine learning called Trillium and said it would license an object detection design and one that could handle some basic training at the edge. Amazon, too, is building a chip for its edge devices and machine learning will certainly have a part to play.
Also on this week’s podcast, Stacey and Kevin cover Intel’s smart glasses, Kevin’s opinions on the Apple HomePod and Google’s new IoT hire. They also answer a listener’s question about using different profiles with the Amazon Echo.
The guest this week is Alexandros Marinos, who is the CEO of Resin.io. He discusses the popular hardware platforms for prototyping, the industrial IoT and an up-and-coming platform that is breaking out because of interest in machine learning. He also talks about the similarities and differences between servers and connected devices as it relates to building software to manage them. You’ll learn that servers are like cattle, not like pets.
Based on the early reviews, Apple’s HomePod sounds amazing. It’s so good according to some that it rivals audio equipment priced nearly three times the $ 349 you’ll pay Apple for a HomePod. And in my testing, I agree: Apple has engineered an excellent sonic experience from a single unit. Plus the microphones are nearly flawless at hearing your voice commands regardless of how loud you’re playing music.
Is the speaker worth $ 349 when you can spend a similar amount on other smart speakers? That’s a difficult question to answer for a few reasons. Generally speaking, if you’re all in on iOS and Apple Music, plus you don’t mind waiting for Siri to get smarter, you’ll be happy with a HomePod. I qualify on the first part of that equation, but not the second. And to be honest, I’m not sure the HomePod sounds that much better than some other speakers that have more smarts.
By that I mean most of the “smarts” in the HomePod are in the sound experience. The device automatically configures itself for optimal sound when you first set it up. And HomePod repeats that algorithmic optimization whenever you move it. That’s smart. Does it really solve a problem though?
Yes, the intelligent configuration is impressive. It’s also easier than the process used on my Sonos One speakers: The Trueplay Tuning requires you to walk around your room as the Sonos app listens to tones from the speakers. This manual effort takes about a minute and, just like the HomePod setup process, it only works on Apple iOS devices.
Here’s the thing though: How often do you physically move speakers that plug into an outlet? Not that often, if at all after the initial setup. While Apple has made this process “magical”, it’s not something you do daily. HomePod will also dynamically adjust music in real time too, although I haven’t heard much of a difference with this feature.
Additionally, I did a bit of a blind listening test with my family and one of my tech-savvy friends, mainly because I didn’t really prefer the HomePod audio over a pair of Sonos One speakers in most cases. That may seem like an unfair comparison because the HomePod is a single unit, while a pair of speakers are obviously two units. So why the comparison from an audio standpoint? Because both setups cost the same: Sonos dropped the price of a Sonos One pair to $ 349 for a limited time.
I set up the listening tests using the same songs in various genres directly from Apple Music and at the same sound levels. More often than not, the Sonos Ones were the preferred option. Note that I’m not saying the Sonos “won” for a specific reason. While the HomePod may technically be the better device for accurate sound reproduction, it’s more important which speakers deliver the sound the listeners prefer. It’s subjective based on taste and hearing capabilities. David Pogue performed a similar blind test on video and nobody chose the HomePod as the overall winner either, further illustrating this subjectiveness.
To my ears, the HomePod is better in the lower, bass frequencies and is impressively good at bouncing sound off walls with its seven tweeters to create an immersive stage. One HomePod is surely better than one Sonos One. Add a second Sonos One though, and the stereo separation is clear, plus the mid-range and high frequencies are more nuanced to me. Again, this is subjective to my ears; I recommend testing any speaker with your preferred music genres.
Unfortunately, most of the “smarts” end there for HomePod and for that you can blame Siri. The best way I can put it is: Siri is fragmented between iOS devices and HomePod. You’d think everything Siri can do on an iPhone or iPad could be done on the HomePod. It’s not even close.
Sure, the HomePod has the basics. Obviously, Siri is super for voice control of specific music or for suggesting playback based on an activity. As I’m writing this review, I asked Siri to “play music for studying” and she was up to the task: I have some easy listening and acoustic hits playing. She knows the weather, the time, can set reminders, and can tell when my soccer team (technically, my English football club) plays next. And of course, she can control any HomeKit device in the home. This all works great.
Want to know your next Calendar appointment or want to create one? Nope. Need to set two timers with Siri? Sorry, she can only handle one at a time. Oh, and although HomePod works for speakerphone calls initiated from your phone, you can’t start a call from HomePod.
Perhaps the most baffling omission though is in regards to HomeKit. In the iOS Home app you can create routines to group different HomeKit devices together and make them do things with a single Siri command. HomePod appears as a device in the Home app but you can’t include the speaker in a routine. I do this with my Google Home by telling it I want Relaxation Mode and it turns the lights on at 25% in my office while also firing up an acoustic playlist on the Sonos One. That can’t be replicated on HomePod, at least not yet.
Apple says that more features such as multi-room audio and stereo pairing of HomePods is coming later this year. I suspect Siri will be improved as well for things like calendar access and the ability to recognize multiple users. The latter is another big omission for me because HomePod is tied to a single iCloud account, meaning even if the calendar features were available, they would only work with my calendar account. My family would be out of luck, unless of course each person had their own HomePod. (That’s not happening.)
Circling back to the beginning, I do think iOS users with Apple Music and HomeKit devices will be thrilled with the sound and home control of HomePod, provided they can wait for Apple to address some of the gaps in Siri’s smarts. Just remember that HomePod only works with Apple Music (for now) and that it doesn’t work at all with Android phones even though it has a Bluetooth 5 radio inside and there’s an Android version of Apple Music. I wouldn’t be surprised if HomePod stays iOS only for a long time, or for good. So you’d better be sure you won’t switch away from iOS if buying a HomePod.
For me (and my ears), a pair of Sonos One speakers sounds very comparable to HomePod at the same price. They also work with dozens of streaming music services and have the more capable Alexa built in now with Google Assistant coming later this year. My HomePod was purchased out of pocket with our site reimbursing me; if I was spending my own money, I’d pass on HomePod for now with a wait-and-see attitude as Apple improves the smarts of its smart speaker.
We’ll keep using the HomePod over time to assess new features and functions as they become available. In the meantime, comment below or call in on our IoT Podcast Listener Hotline at 512-623-7424 if you have HomePod questions.
Did you know your television is watching you? Specifically, that most smart TVs are sending data off to their makers and in certain cases, to marketers. Consumer Reports showcased the security flaws and the lack of privacy inherent in connected TV in a report last week, while over at GizmodoKashmir Hill has a new article out about privacy in the smart home that puts a big focus on televisions.
It’s no secret that internet-connected TVs share data with others, nor is it remarkable that most TVs available today are smart. That’s what allows you to watch Netflix, YouTube, or Amazon Prime shows. But the rest of our appliances are also going the way of TV. Samsung and Kenmore both say that, going forward, all of their appliances will have some kind of connectivity built into them.
And for many, the features enabled by connected devices will mostly outweigh the fears of data surveillance. I’m not talking about connected light bulbs and home automation here, but about adding truly innovative and helpful features to once-dumb appliances, letting them become truly smart.
An example of this is a washing machine that can tell how dirty your clothes are and select the proper cycle. Or a fridge that can offer you a remote camera feed to the inside so you can see what’s on the shelf. Maybe the fridge could reorder your water filter when it’s getting old. Even better, maybe that same filter could report back on the purity of the water to environmental agencies and consumers as a way to ensure public health.
Smarter products will have to be connected in order to create information exchanges that benefit the consumer, the manufacturer, and maybe even society. However, the industry so far is screwing this up with an ineptitude driven by greed, short-term thinking, and a desire to act first and beg forgiveness later.
This is emblematic of the culture built up over the last two decades in technology, where we took the internet and used it to turn users into the product. The current backlash against Silicon Valley companies is a reaction to this exchange of personal data for services. Especially as the services became more about keeping the person engaged to the exclusion of their well-being or the well-being of society.
This may sound like hippie dippie stuff, but there is a direct link from Google and Facebook’s behavior to the privacy concerns that people have with regard to connected devices. That those concerns are completely justified only makes it worse.
I’ve spent years trying to tell the industry and the government that privacy matters. Not just because it’s a basic right, but because if you respect people’s privacy and offer them agency over controlling their data, they are more likely to buy the product. And if you offer them a compelling reason to share their data while still offering them some control, you actually build a model where the data you collect has to benefit the user or the larger society.
We are starting to see some momentum on this front, and I am hopeful that 2018 will be a turning point in the U.S. The General Data Protection Regulation in the EU has already established a framework for how to establish data privacy as a human right. What’s even more promising is that many of the regulations in the GDPR are impossible or difficult to implement today, and the EU realizes that.
The hope is that the EU will guide technologists in developing tools that match the regulatory framework while the regulatory stick offer will offer an incentive for companies to make a market to develop the tools required to meet the law. Meanwhile, here in the U.S., technologists are increasingly asking themselves how to get and use data responsibly.
While this entire essay is focused on the importance of managing user privacy and the intentional gathering and sharing of consumer data, security is also related to the topic. Specifically, what happens to consumer data when security is breached. As it stands, consumers are worried both about a loss of their privacy to companies, but also to hackers as part of the all-too-often security breaches.
Until the tech companies get their priorities in order and the government steps up with rules that give consumers some control over their information, I believe the promise of the smart home will never take off, because consumers won’t trust it.
As inspiring as the phrase business transformation is, I’ve decided that when it comes to industrial or enterprise IoT, it’s better to start small. Most executives by now are well aware that you should begin with a use case, but what’s become more clear as time has passed and projects have failed is that maybe business transformation shouldn’t be your first goal.
Peter Zornio, chief technology officer at Emerson Automation Solutions, says that in his experience, the operations guys in building systems or in a plant want a use case and an ROI, while the IT shops tend to want to install a platform so folks in the business can build their own applications on top of it.
“Tangible ROIs that are easy to see are great,” Zornio says. “Operational guys love that because they have to justify their spend, while the IT guys want to think big. These are the guys that 15 years ago convinced everyone to spend hundreds of millions on ERP systems.”
Zornio isn’t bashing ERP systems, but if you ask ERP buyers if that money was well spent, many of them wouldn’t really know. Which is why Zornio is a big fan of metrics when discussing IoT projects.
He’s not alone. Jason Shepherd, a senior director and IoT CTO at Dell, says, “Too many IoT projects start as science projects (e.g., “Wouldn’t that be neat?”) with no clear metrics for success.” You know what’s really hard to measure? Business transformation.
So if measurement is the key, how should you think about that? In some situations, a use case and the subsequent savings are crystal clear. For example, if you automate data collection that normally requires an employee, calculating the savings is easy.
But Zornio says other use cases, such as ensuring reliability, are more difficult. First you have to come up with the number of times a particular part or machine fails, then you need to figure out the cost to the production process or the team. You also have to factor in the cost and time it takes to make those repairs. Replacing a part that is commonly in inventory vs. replacing something that might have to be ordered will factor into those costs.
Those kinds of calculations are more subjective than calculating the cost of replacing a worker. You could debate how often equipment fails. Or how much it costs when it does fail, depending on what a company values. For example, downtime in one part of the plant might be relatively unimportant because there’s a backup or low demand during certain times of the year. So it’s always better to search for the obvious. Sometimes, the flamingly obvious.
“We had a customer come to us about monitoring pumps. There, the risk wasn’t downtime, but that when one of the pumps failed it tended to catch on fire,” says Zornio. “In that case, the ROI wasn’t about money saved as much as it was about deciding how valuable it was to the organization to avoid fires in their factory.” (That entire conversation has me thinking that an enterprising IoT systems integrator should scour the trade press for industrial disasters to find their next sales prospect.)
Assessing IoT projects’ value isn’t just useful for the companies buying into connected sensors or products. It’s also important for companies trying to build solutions for industrial and enterprise IoT.
That platform mentality is a common one in Silicon Valley, but it’s hard to sell. Especially if you need a deep understanding of specific industry data around costs and functioning of equipment. That’s why many of the big companies are teaming up with those in specialized verticals to pitch their platforms or services.
But again, it appears that success today is found most often in the smaller projects as opposed to the business transformations. Shepherd advises that when choosing a project to ensure that the use case is relatively straightforward so the company can get a “quick win.”
“A quick win can grow into more advanced benefits, but don’t try and start with too much. For example, start with basic monitoring for visibility and then add analytics,” he says.
We’ll talk more about what this means in future issues of the newsletter along with the challenges associated with making sure that your employees don’t sabotage your business goals—or the eventual business transformation itself.