Bit quiet of late...

Probably spent more time than we'd like on testing and debugging the new version of Cortex on a number of existing and new installations. There should hopefully be an announcement very soon on this.

As for the latter, no tougher than usual . In some ways the hype helps and in others it doesn't. I presented an item on a particularly useful application of our technology (which would be pretty hard to match by any other) at a conference a couple of weeks ago. It was enlightening to hear the constant misuse of terminology just about by everyone. Something 'smart' was anything that was 'connected' and 'connected' meant to a cloud service. Followed by a 'smart' home being defined as just about anything which had a handful of such 'smart' devices. This then was followed by people providing various and, as you might imagine, wildly differing statistics on aspects such as the adoption rate of 'smart' homes in various countries, peoples' perceptions of smart devices, the value of the market for smart homes, and so on. Of course nothing new in all of this but it did kind of bring it home to me how some of the plot had been lost under the hype and also perhaps why people often become disgruntled by the promises of the so called smart home. To say nothing of some claims which suggested that people had found a universe with alternative physics .. .

Idratek / Cortex certainly offers capabilities way beyond what MS / Amazon / Apple seem to have so-far imagined ... am a fan of them all ...

we were at a (Broadband) workshop the other day, and (our particular set-up having become locally fabled for it's advanced capabilities) first question we were asked was if we had AV (TV, HiFi, etc) integrated in ... what could I say ?

NB: do recall Paul's integration of Squeezebox ...

And here's another anecdote ... been trying to help source some electric binds recently and was really a bit surprised how difficult it was to get some suppliers to provide a quote for a simple 230v motorised set up with up/down/common connections, which we could then control via a relay module - having already routed power, network cable and back boxes to convenient locations for 6. Most seemed to start off suggesting battery and radio based products and then went to radio based 230V driven units but started to suck air when we said we needed just the plain motor (which in fact is cheaper than other two options). Couldn't understand why we wanted to use 'old fashioned' technology. I asked them how long they expected the batteries to last. Most answers were between 6-12 months ('a bit less at the beginning because people tended to play with them more'). So, that's six blinds with fairly hefty dimensions each probably being operated more often than manual control due to automation, with a battery life expectancy of 12 months each (I doubt). Meaning you could be changing/charging batteries every 2 months, let alone the possibility of them conking out when you were away. Then there was the question of integration.. yes it might be possible via this gateway and that cloud service and IFTTT... well I suppose better than hacking a handset, or is it?

I understand fully well that the reason is that the market for retrofitted blinds is huge, so everyone is focussing their resources on this, but for those who are able to do the job properly I think it is going backward not forward in terms of the final user experience.

did you look at Somfy ...



being what we planned to use for a couple of windows, off a DRB relay ...

Chris

PS: Googled for possible source ...

Karam,

Glad to hear things are still doing well and the Idratek team are bringing more things to the market. "Smart" as you suggest seems to now mean connected so you can operate with your smartphone, but to us long term HA enthusiasts this doesn't seem very smart at all. A true smart home sorts itself out rather than require manual input from a phone.

New technology and cost reduction should've made HA easier for the masses but that doesn't seem to be the case. Instead every manufacturer wants to develop his own standard on his own web server with no integration to anything else. Machine learning and AI have the potential to make things like natural language interpretation much more accessible, the compute power and relative low cost of "cloud" is something that wouldn't have been imagined five years ago, but is this going to be realised?

One thing that does seem to be improving is the capability of IC at the same time as costs going down, guess we have to thank China for that. Maybe, just maybe I'll get my wish and see some energy reporting and switching at the appliance level in the not to distant future in an Idratek module (for now the Sonoff PoW with a custom image integrated into Cortex is as close as I've got)

Paul

Still adore my various Squeezebox. Gutted they are no more and once the units I have die it will be the end. My rave about Sonos but they are expensive, less capable, less flexible and don't integrate

thinking of integration beyond ... Kevin & xAP comes to mind ...

but xAp seems really quiet, these days - some sort of interface with Cortex was developed, but never did get to understand it ...

Yes indeed, just about everyone seemed to mention or use their motors and it was the Sonesse 40 WT I had in mind

It is far easier to market a 'single point product'. For example it is not difficult to describe a thermostat to your potential customers and at the same time they are really VERY forgiving to any downtime or so called 'intelligent learning' errors. It is also easier to sell something like smart phone controlled lighting. Again easy to understand function, so if easy enough and cheap enough to install then people will no doubt be tempted but perhaps eventually get disillusioned when they realise its a bit of a pain having to navigate your smart phone just to turn on your room lights. Even the use of a voice assistant to do this isn't so clever - lost count of the number of times I've been asked if I can turn my lights on (and of course off) with Alexa, to which I've replied - have you actually thought how much of a pain that would be to have to do routinely instead of automated lights? Of course I'm not saying it wouldn't be useful as an override method, but then the concept of fully automated lighting is itself not actually so widespread - most people have bad experiences of motion only based lighting automation, and automated lighting tends to actually be perceived as 'scenes' or perhaps sunset/sunrise schedules for coarse lighting.

What our own hypothesis has always been - is that to get any useful kind of HA you have to take a holistic approach and you have to be prepared to use quite a lot of devices, with good integration being a key factor. So although the technology for building and connecting devices has got cheaper and more powerful the philosophy of most manufacturers has not changed hugely, and quite frankly from a business point of view why would you want to expose yourselves to the headaches of supporting a complex overall product when it is far easier to sell single point gadgets? Probably the time will come when this aspect can be more easily distributed, but I think we even have to get to the point where people grasp the differences between stateless and stateful structures.

Yo .. I’m still around and a huge proportion of my home chats away on xAP still, including Cortex. It’s just so easy to add xAP support to little devices and, as you mentioned them.. the xAP support on SqueezeBox devices is excellent. The loose (broadcast) ‘udp’ coupling is very powerful still but it’s a tough sell and so time consuming that I keep my head down nowadays.

As things have evolved Node-RED and MQTT are very much in the frame for me now and JSON payloads even transported within xAP. It was an easy bit of code to transition by pushing every BSC state change from xAP onto MQTT and vicaversa and hence get Cortex reporting status via MQTT. Although I personally haven’t implemented the reverse ... updating devices in Cortex via MQTT ... it would seem very do-able either via xAP or the API.

What’s great about this ‘middleware’ approach is that it offers so many paths to interconnect A and B either directly ... or via C and that just makes everything possible in HA. It can sometimes get a bit confusing as to quite ‘what’ originated a change and you can get loops if you’re not vigilant but it’s very powerful. You can basically integrate any device supported by any HA app into any other. As this now encompasses so many cloud based services offered by the Amazon / Google assistants there are amazing integration opportunities. IFTTT, Conrad Connect and Stringify are also great in this respect, bringing so many software linkages into the mix. It just makes me smile nowadays when I get something ‘talking and walking’, which is basically the fun of HA for me.

My main concern with current home automation product directions though is Cloud / Internet dependence to do almost anything. In some cases when you press a switch to turn a light on it fails if your Internet is down as does any scheduling / logic. Not to mention the inherent delays. You MUST have local control in a HA System. Cortex, along with it’s directly attached hardware and node<>node fallback intelligence still excels in that respect.

K

>stateless and stateful ...

had to look that up ... best I found pointed to stateful being in effect living, intelligent ... and stateless being dumb, effectively moronic, in its responses ...

Kevin, you'll be pleased to know that I mentioned xAP at the afore-mentioned conference just a couple of weeks ago - in the context of interoperability mechanisms

Chris, just my terminology... Distinguishing between systems whose behaviour depends on a state model (i.e dependent on the states of multiple objects and signals at the time) as opposed to systems which simply act on a narrow set of immediate local data. For example when you press a button to change the state of a light in Cortex the resulting action and how it is achieved depends on what caused the light to be in a particular state in the first place. In a stateless system the button simply toggles the light without being concerned about any consequences on the goal of automation (e.g under what conditions automation should subsequently affect that light). Similarly for occupancy inference - the system has to have a picture of the whole house including multiple sensor types and historical information, not simply work on the immediate signal of a motion sensor. A bit like the human brain maintains a holistic model of the body and the environment rather than simply reacting to singular signals, though of course simpler reactions possible via reflexes.

so, how long before people realise the difference ... what could be done to tip the balance in their understanding ... many people are not yet enamoured of HA, and maybe this is why - albeit they couldn't articulate it ...

I suppose ultimately its about the user experience - what the technology does on a day to day basis and what flexibility it gives you beyond that. Whether a stateful or stateless approach is better or whether a clear distinction can be seen in various technologies is probably something for the enthusiasts to debate.

However I know only too well from personal experience how difficult it is to get across the differences between the 'feel' of a home using one kind of technology relative to another. Probably needs a decent video, or I'd go as far as saying a serious documentary - not one of those that simply sets out to glamorize, eccentrify or nerdifiy the technology (hope I'm not adding new entries into the dictionary).

I tend to use our case study 7 as a good example, because it illustrates usefullness over glamorous gadgetry and co-existence with a user who has no affinity for technology. Of course if I want to impress someone with scalability then its over to you Chris