Hub or Hub- less smart home? what is the real difference?

For anyone that’s seen me present (meet-ups, industry events, accelerators, investment pitches) you’d know I always carry around one of the first prototype blue boxes and hold it on stage when talking: clearly nCube Home is a Hub-based Smart Home solution.  But there are also lots of app-only solutions around too.  

When I present I never really think of the difference in terms of hub and hub-less but recently when I spoke at the Brighton IoT Meet-up and also in several recent partner meetings I’ve received feedback on how useful it is to know why a hub makes so much more sense - so hence worth a blog article in itself.

Our focus is to make a great design-led, user experience; to focus on home privacy; and to work with a growing list of smart home devices.  In the world of smart home, creating a really good user experience across a wide range of devices is hard work so we’re really pleased with what we’ve achieved so far.

App-only solutions can similarly work on delivering great user experiences sitting on top of the integration point of say Philips Hue lights - for example there’s some great innovation going on with in-app purchase of clever scenes like a fireplace mode.  There are two main methods to that connectivity - either using the integration point of the device in the home or using the integration via the servers provided by that device.

And there’s the first problem - if your app has to use the in-home connection then the moment you’re away you lose the connection.  And if you connect through the server/cloud integration then you probably lose privacy control and you start adding in latency.

So by dint of being a hub and our commitment to privacy we solve those two issues by default as we always aim to connect the hub to the local integration point (The Nest thermostats is the only exception at the moment) and the hub obviously carries on running when you leave the home - the nCube app has no logic or service layer running in it and neither do our servers.  In fact at one recent event a leading app-only solution enquired about hosting their app on our hub because of these issues and we’re now working with them to make that happen.

Next is the range of devices.  Probably only 25% of the devices we’ve integrated are directly accessible over your home wifi - things like Playbulb, Parrot Plant Monitor (both Bluetooth), Aeon Labs, Everspring and many others (Z-Wave radio).  So you’re very unlikely to see a large range of things supported by the app-only solutions.  Some Bluetooth devices might be accessible to app-only solutions but only when your mobile is at home and certainly none that are Z-Wave, Zigbee, Thread and other non-IP centric radio layers.

The nCube hub acts as the gateway, or bridge, to each of those different technologies so you don’t have to start checking compatibility lists.  So far in the past 5 years no single radio/connection method is winning - we’re just not seeing the sort of race that happened between High Definition DVDs and before that Betamax and VHS formats.  If anything there are more versions coming out - Bluetooth 5 is on it’s way, there will be a low-power mesh version of wifi, smart meters are rolling out over the next few years with their own distinct flavour of zigbee, etc.


And thirdly, app-only solutions rely on devices having cloud services to integrate to.  Some also use other cloud rules/automation solutions.  As that happens your private data starts being tracked through several cloud servers meaning both privacy concerns and introducing latency as your thermostat control routes via cloud servers.  Not to mention security risks.

Several times a year the various leading thermostat providers experience server issues.  Layer that on top of the stability of home internet connections.  Then add in that an app-only solution may be connected to quite a few different cloud services and the likelihood that something is going to fail starts to become very high - perhaps even experienced in some form or other more than once a month (e.g. I know of some heating solutions failing on average twice a year; if I have four different products and if we assume home internet connections have an issue a couple of times a year then that’s 12 issues in total per year in total).

So to summarise: 
1) Device integration locally or via cloud-servers - adds privacy, security, latency risks. 
2) Supported device range - app-only solutions may support less than 25% of the sorts of products around. 
Offline working - a home with 5+ devices using an app-only solution could experience over a dozen issues a year. 
3) Privacy - if there’s no hub there’s no clarity on how private data is and how that’s managed.

I'd love to hear from you about your opinions on the differences between hub and hub-less Smart Home systems. Send an email via or find me on Twitter @rewllen


  • Nigel Walsh

    Nice summary. My personal challenge with this is the battle for the hubs. It seems with every new cool thing to try I end up adding another hub. I’ve tried to break it down to :

    1. Controller interfaces (eg app, thermostat, voice (Alexa etc) and web),
    2. Integration kit (I put HomeKit, Smart Things or IFTT in here)
    3. Physical Devices (all the IoT or end point stuff)

    which I think it quite similar to what you are doing here. If I look at my house alone, and I appreciate I like to try these things out so don’t judge me! HomeKit compatible stuff is marked as HK. I want to add more stuff, but until I solve some of this I fear I’m just going to be adding to the mess. For example, on lights/routines – I can do this in Phillips, Alexa or SmartThings. Heres a few..

    Heatmiser NEO – Hot Water / Heating (HK with Hub Upgrade)
    NEST – Smoke & Carbon Dioxide
    Phillips Hue – Lights, Lamps etc (HK)
    Belkin WEMO – Netcam Cameras (ditched these), WEMO Switches, (have ditched the bulbs, moved to Philips) – Very poor hub connectivity.
    Ubiquiti – Unifi Long Range Networking
    Samsung SmartThings – Plugs, Door Sensor, Water Sensor
    SONOS – Multi Room Music etc

    IFTT – for linking the Services together in the cloud
    Apple HomeKit (HK) – Some of these now available on IOS 10.

    Leakbot – Water Leak Detection
    NEOS – Facial recognition, Moisture, Movement, Doors
    Alexa – Amazon Echo
    Dahua – IP Cameras

    My litmus test though however is Mrs W. if its easy to use, and passes my travel test (If I go away a) does it still work (too many examples of her calling me to turn a light on when abroad or the heating coming on at strange times due to my phone being in a new timezone) and b) can she operate reliability and easily!.

    The last part here is it needs to be affordable. This is already a considerable investment for something not fully integrated and ‘seamless’. There will always be specialists for each area with high end industry leading products, but coming back to the rest of us – I feel there needs to be an affordable solution for the everyday or lifestyle user.

    Out of interest, where do you put things like or Control4 in this – would you have them as alternatives to nCube or firm competitors?

    Rather than all being connected to everything, It needs a single easy (beautiful) interface that allows adding of hubs seamlessly and new kit etc that then doesn’t create new conflicts in the system. I think what you are saying here is nCube solves some (or all of this) going forward by creating a master/slave approach.

    One thing is for sure, these will only get smarter and more connected over the years.

  • David Hovey

    Great work Phil, well written and the situation could not be more as you explain. I think also added to that are the many non-radio automation issues – what about the hard-wire stuff, a huge amount of legacy gear whether it be boilers, thermostats, detectors/reed switches, the works to connect in – any thoughts on the even less modern ‘gadgets’ in the home?


Leave a comment

Please note, comments must be approved before they are published