Blog Post

Belkin looks at the smart home and doesn’t see a place for hubs

There are more than a dozen smart home hubs on the market. They range from $50 for the Wink home hub sold at Home Depot to the $300 Revolv product. The idea behind each of these boxes is that the many different radio standards available in the smart home market as well as the need for software to make the experience easy for consumers necessitates some type of central box inside the home.

But Ohad Zeira (pictured above), who manages the WeMo ecosystem of connected home devices for [company]Belkin[/company], doesn’t agree. The company, which did end up making a hub product for Staples Connect as part of a holdover deal made when [company]Cisco[/company] owned the Linksys brand, doesn’t plan on making any more hubs. “You need a hub for a unified and interoperable home today, Zeira said. “But consumers don’t want apps or a set of rules and as soon as a standard becomes real the hub will cease to exist.”

He looks at a hub as both a high cost of entry for consumers to add intelligence to their home, and also a potential source of unnecessary complication in the form of setting up the device and then using the software to program their devices. “Hubs are designed for early adopters and the industry is trying to craft this idealized vision, but solutions that appeal to early adopters aren’t the ones that will cross the chasm to the mainstream.”

We Mo app for Android

Belkin is using Wi-Fi for its line of connected outlets, cameras, light bulbs and other devices, and Zeira believes it could even become the most power efficient standard as low-power Wi-Fi (802.11ah) eventually hits the market. Until then, WeMo is pushing interoperability for early adopters through its integration with If This Then That, where people can link their WeMo products to other devices and build complicated programming rules.

Zeira also thinks the intelligence people might need to program their homes should eventually live inside the router (a device Belkin currently sells a lot of given its ownership of the Linksys brand). Belkin is building that intelligence into routers sold to the end-consumer. And while many consumers get a router from their ISP, modem fees and issues with performance drive many consumers to buy their own. According to IDC, around 60 percent of routers shipped in the U.S. are sold through retail channels.

So for the mainstream consumer, the smart home hub features will eventually be folded into a router with companies like Belkin or others providing a curated set of devices that help the consumer make their lives immediately better. More advanced users can tie systems together via [company]IFTTT[/company] or other services. Plus as part of that curation effort, Zeira plans to continue integrating WeMo into consumer products like the connected slow cooker. Belkin signed a deal with Jarden Corp., which owns the CrockPot and Mr. Coffee brands. A connected Mr. Coffee coffee maker is expected soon.

“No one ever says to me, ‘I feel really disconnected from my home,'” Zeira said. “People have jobs to do and they need devices to do it. That’s what we should provide.”

13 Responses to “Belkin looks at the smart home and doesn’t see a place for hubs”

  1. Konstantin Bogun

    I surprised of your point of view, or your advisers that contributed this evaluations. Really, Router + Smart Hub = new decision for all wirelesss technologies in one bag. It is very simple and clear, but its first article which covered this thought.

  2. Yuval Lachman

    Absolutely agree with the concept that ” for the mainstream consumer, the smart home hub features will eventually be folded into a router”. The routers should also take care for the secure remote access to all of the THINGS connected to it. I think we will start seeing more and more routers adopting the Open WRT, DD-WRT or others open source operating systems that allows establishing a secure data encrypted channel like SSH or Open VPN which support the port forwarding to be done at the cloud instead of inside the router. This actually will allow the mainstream consumer to get a seamless plug & play feel for the connectivity between their smart products and their router home gateway. No need to do all the hard and complicated work of setting and configuring the routers…

    • Joe Schmelzer

      Interesting. I see both advantages and disadvantages of that architecture…

      Example, one of the benefits of a localized architecture is the avoidance of the WAN connection, and all that entails. (Like the NEED for a WAN connection…)

  3. Martin T. Focazio

    I feel old and grumpy because I think that the whole IOT world is a security disaster, and hubs represent at least a line of defense that can be implemented to mitigate the risks of directly-connected devices. I’ve already had my dose of security reality with directly connected to the internet IOT devices via the Belkin WeMo line – it had deep and easily exploited security flaws that they ignored and then were slow to fix. While I really enjoyed being able to water my garden while on vacation by turning on an irrigation pump, I did not enjoy a device that was effectively opening up a hole into my home network. Hub functionality in an upgradeable router that handles both TCP IP and IOT device communication protocols is more likely. Eventually, as we move to IP v6 (well, maybe some day) we’ll get to a point where each connected home will need have the rough equivalent of a windows domain controller and it will be sort of like part of the HVAC systems – just another appliance, this one managing the home network services and devices. You won’t mess with it often, it will sit there blinking most of the time, but like a furnace, it plays a huge part in your home. This home gateway (not a “hub”) would serve as the firewall for all the stupid IOT devices that will never be able to defend themselves from outside attacks and will never, ever be patched as well as serving as a way of managing the users and other smarter devices hanging off the network. But right now? Right now IOT, hubs or no hubs, is roughly like “Going Online” in 1990. You picked a service (AOL, Compuserve, GEnie, Prodigy) and you were stuck with their software, their services and their cost models. Until I can pick an iOT light switch based on HOW IT LOOKS not how it works it’s going to be a nerd market only.

  4. First of all I’d like to know if he’s even tried his own products. WeMo is a terrible user experience, terrible software and doesn’t work well when you’re not on the same network.

    He’s clearly not taken the cost difference between Wi-Fi and ZigBee into account, Wi-Fi requires both external RAM and Flash memory which adds a lot of cost for what should be very affordable devices. A good ZigBee solution doesn’t require much extra in terms of components and you can technically do the entire ZigBee part of a sensor/device for less than $2 today, whereas the equivalent with Wi-Fi would be around $5, somewhat depending on how much memory is needed. Give it another 6-12 months and ZigBee will be sub $1, but Wi-Fi will still be $5+.

    Sure, hubs are kind of annoying, but there are solutions. The long time coming Almond+ router has both Z-Wave and ZigBee built in and it reduces the clutter hubs can add.

    ZigBee and Z-Wave also have the advantage of of being mesh networks, whereas Wi-Fi does not when it comes to home automation. 802.11ah is a terrible “standard” due to it being sub 1GHz which means in each region/country it’ll need a different frequency and it’ll range from 700-900MHz which means a single product design isn’t going to work.

    It’s also funny that he claims that hubs are bad, especially as Belkin just launched their own hub for their ZigBee bulbs. Not hard to see why, as ZigBee takes up a lot less space inside something like a bulb, so you can make a smaller bulb. Oh and it’s cheaper as well…

    • Joe Schmelzer

      I find the material cost argument to be pretty thin, short term… OK, so you pay (in material cost) $4 * n devices, one time, in your home… So what? Let’s say I’m a mega user and have 50 devices. Wow, I just spent an extra $200 to have them all be Wi-Fi. (By the way, my 50 devices cost me, say, an average of $50/ea, for the sake of argument, so, $2500.) I spent $2,700 but get to keep my routers from CISCO/BELKIN/etc. instead of – ahem – who was it? Almond+?

      Belkin does a pretty decent job at knowing what consumers like. I’m betting a person buying 50 devices will pay $200 to have them connect to Wi-Fi, and existing infrastructure, rather than have to install new protocols/hubs/etc.

      And, have you ever worked with Zigbee/Zwave equipment directly? Other than HUE, I haven’t found a passing UE yet. None are good.

      Mesh network stuff is useful/cool, but it’s almost a circular argument. Mesh networking obviates the need for direct communications. But if my thing is already talking to the Internet, I’ll manage it from a cloud/app anyway, and my mesh-like features can all be managed there… (see Quirky) Other example, HUE lights. Not sure if it’s Zwave or Zigbee, but one of them. Anyway, mesh is irrelevant in this use case because all management is done through mobile app – with HUE hub (physically) plugged into broadband router…

  5. Joe Schmelzer

    Ohad’s logic is sound. A specialized hub is another piece of equipment a user doesn’t want to have to purchase. [I love my Philips HUE lights, but when I opened the box and saw that hub, even though I knew it was in there, my heart sank a little. Happily the user experience is WAY BETTER than any other light out there…so I’m OK with it for now.]

    However, getting a little technical here, the hub currently plays an important and necessary role. IoT devices – whatever they are – want to be connected to the Internet. To do so, they need addressing, which is currently done through TCP/IP. Running TCP/IP software, and having the computing power (cpu) to do that, is not zero dollars, size, power, etc. In fact, it’s pretty costly, particularly when multiplied by 50 billion.

    The hub allows for those edge devices to NOT have to support TCP/IP and the Internet directly. Makes for lower cost, lower power, no TCP/IP stack, etc. But you need a hub.

    Today, when you have just a few devices connecting in the home, meh, who cares if I have I spend a few extra bucks on each one to have them connect directly to my W-Fi router? I prefer that usability, for sure. But again, when that is multiplied by 50 billion the economics are different.

    Of course, that 50 billion scale might bring the cost down to where it’s negligible… Anyway, I agree with Belkin’s philosophy that hubs are annoying, stuff should connect directly. I just wouldn’t wait around for the golden standard. Many of us have been doing that for a long time. I started waiting in my 20s, now I’m 45.

  6. Adam E Nagy

    Based on the title I thought he would say that eventually any device will be able to communicate with any other device without the need of something else in between: a router or a hub. But nope :)

  7. Person says “There’s no need for hubs!” then says “Router will be the hub!”

    Sigh. Regardless, a hub will be important unless these IoT devices are going to be handling security on their own, and that’s going to be cost/power-prohibitive for some devices.

    • Adam E Nagy

      Based on the title I though he would say that eventually any device will be able to communicate with any other device without the need of something else in between: a router or a hub. But nope :)