Posted by: @knukes

↑

@cathoderay Exactly. I imagine at some point artificial intelligence and machine learning will take the calculations and weather compensation away from installers so you can just plug and play. That way, the system will optimise itself.

I am working on that, the first prototypes have been tested at a university on a unit that produces heating and hot water which comprises of a box the same scize as a combi boiler and a small heat pump outside. It is compleatly autonimous, commissions it'self and runs it'self continually adapting to the enviromet and heating system to produce the best possible comfrot levels and efficiency. All the installer needs to do is connect it to the internet, no more difficult than connecting you laptop.

I am not too far from reliese. I am currently seeking funding or investment for the next and final stage of developement.

@heacol That sounds fascinating and best of luck with it. Have you approached any of the manufacturers for funding?

@knukes 🤣 🤣 🤣 🤣 Yes, they all say it cannot be done. We have a report from the university saying it does work but we have a long way to get to commercialisation.

Posted by: @heacol

↑

@knukes 🤣 🤣 🤣 🤣 Yes, they all say it cannot be done. We have a report from the university saying it does work but we have a long way to get to commercialisation.

Industrial control systems can perform more complex operations, but of course come at a price. Though having said that, the industrial controller I have installed on our heating system can keep the indoor temperature within +/- 0.1C of the desired temperature, and only cost a few hundred pounds.

@derek-m but only you can make it work, therefor useless on a wider scale.

Posted by: @heacol

↑

@derek-m but only you can make it work, therefor useless on a wider scale.

That is true to a certain extent, but self tuning controllers have been available for many years.

When I am setting up control systems, I spend some time initially just monitoring the response times, to get a feel for what are the important factors that need to be taken into account.

The major problem that I find with heat pump controllers, is that one cannot interrogate the software to see what it is doing and how it is responding. There is also the complexity that no two homes are the same, so the overall system design can be even more important. Having the best control system in the World will not perform very well, if the equipment that it is controlling is not up to the job. We see this time and time again on the forum, where the problem is not the heat pump or the controller, but that the system has been poorly designed and the controller incorrectly configured.

Posted by: @heacol

↑

Posted by: @knukes

↑

@cathoderay Exactly. I imagine at some point artificial intelligence and machine learning will take the calculations and weather compensation away from installers so you can just plug and play. That way, the system will optimise itself.

I am working on that, the first prototypes have been tested at a university on a unit that produces heating and hot water which comprises of a box the same scize as a combi boiler and a small heat pump outside. It is compleatly autonimous, commissions it'self and runs it'self continually adapting to the enviromet and heating system to produce the best possible comfrot levels and efficiency. All the installer needs to do is connect it to the internet, no more difficult than connecting you laptop.

I am not too far from reliese. I am currently seeking funding or investment for the next and final stage of developement.

 

I wish you all the best with your endevour, since I fully support anything that improves systems for the consumer.

If you feel that I can be of any help, then please feel free to PM me.

Posted by: @knukes

↑

I imagine at some point artificial intelligence and machine learning will take the calculations and weather compensation away from installers so you can just plug and play.

Posted by: @heacol

↑

All the installer needs to do is connect it to the internet, no more difficult than connecting you laptop.

And you will own nothing, and you will be happy, to quote from another forum. I have to agree with @derek-m, in that even the most sophisticated heating control system is all said and done a feedback system, possibly even an adaptive feedback system. Such systems can be extremely sophisticated (most of your body runs on feedback systems, for example) but they don't involve intelligence, artificial or otherwise, and it is difficult to see what machine learning, basically observing patterns, can add to heat pump controls beyond what we mere humans can achieve by observing patterns.

For example, I can observe this morning one of the problems with weather compensation. The problem is the compensation system only has one curve. The downstairs rooms are all significantly below design temps, by about 3 to 4 degrees, brought about by the cold snap and a heat pump that can't quite cope. But the control system (weather compensation) has picked up the fact the outside temps are milder this morning, by about 5 degrees, so it has lowered the flow temperature. The system is missing a feedback loop, one that tells it that the house is colder than it should be, and it needs to adjust the weather curve upwards to deal with this. Of course, I can do this manually, but adding that extra feedback loop would make the system 'smarter'. But it's not AI, and no machine learning was used in the above analysis.

It is quite enlightening to ask where AI and machine learning have made a real difference in the real world. I for one struggle to find any salient examples. No doubt someone can enlighten me (not meant in any sarky way, I'd really be interested to learn, in a human learning sort of way).        

@cathoderay You’d be surprised at how often machine learning is now being used in real life. Sectors including manufacturing, healthcare, financial services and others are using it to find patterns in the data and develop new solutions based on that. 

I’ve been involved in quite a few projects myself implementing it. 

While I’m not an expert on algorithms, I can’t see why AI wouldn’t be able to see patterns in the data in your house and outside (assuming you have the right sensors), and make adjustments based on that to give you consistent temperature and efficiency. 

Posted by: @knukes

↑

able to see patterns in the data in your house and outside (assuming you have the right sensors), and make adjustments based on that to give you consistent temperature and efficiency. 

What I am trying to get at is this is what we normally call a feedback system, and they have existed for millions of years, eg biological systems. As I said earlier, just about everything 'automatic' in your body is controlled by sometimes very sophisticated feedback loops. I do know that google and facebot use 'AI' to place intrusive and often of zero interest ads on pages in my browser, eg I read an article on garden gnomes, then I get plagued by ads for garden gnomes, but this is still at heart a feedback system, signal is me reading the article, response is to feed me gnome ads. I can't see the intelligance in that, it is a dumb system 'learning' my from behaviour and acting out a dumb response.

I really interested to know more about AI and ML, because at the moment I don't 'get it'. It all feels rather like a marketing tool to me, you brand up your  new air fryer and having AI, and then reach out on social media to share your passion for your revolutionary AI controlled air fryer.

Where it might get a bit scary is when you forget to tell the buyers of your new AI air fryer that it is connected to the internet (it's part of the IoT, or internet of things) and that now means the air fryer manufacturer now knows rather more about you cooking habits than you would perhaps like.

@editor - I know this is going off topic - or is it? Isn't this (where AHSPs may or may not be heading) something people new to the sight might want to pick up on? ASHPs are a prime candidate for the IoT, and people may want to know that before they leap in and buy one.    

@cathoderay My Enphase microinverters and control gear plus Tesla Gateway are part of the IOT family, as are my Amazon Echo units, our washing machine and a few other things - but I won’t lose any sleep over it I don’t think. AI is an all encompassing term I feel, definitely artificial as it has been created by mankind but intelligence - well as they say, the jury may still be out on that! Regards, Toodles