Also I think the system should be automated… like smart fridges… that use sophisticated algorithms to learn what we like at different times. this should then be the default, with the three buttons for adjustment.

just my 2c after a glass of red

Multiple heat ‘outlets’ even if they’re drawing from a central source as in Somnambulian Pirates’ idea would be very useful where there are people with different body temperatures, or perceptions of temperature (i.e. always, as Mayo’s project addresses) and where different parts of the room have different air temperatures due to cooling losses (i.e. everywhere, until thermal equilibrium is reached throughout the entire system – in the case of Crosbie’s wall, that would mean the entire outside world too!).

Dave’s example of the car heater controls is, indeed probably where the prevalent mental model comes from – and it does intuitively make sense from a user’s point of view.

Some flats/offices have forced hot-air heating systems, where (I presume) there is a central furnace / other heater with fans that pump the air to every room, at some (centrally determined) temperature, and the vents on the outlets in individual rooms can be shut off as required to reduce the outflow of hot air (and thus control the temperature locally). The flat I live in now once had this (there are lots of redundant vents around) but it’s no longer in use.

This, then is surely a real type of central heating which does accord with the valve model (though I guess only up to the point where all the vents are open and the air coming out is at the maximum temperature set by the landlord or whoever.

Rich’s point about the sort of products that people actually buy, even though better alternatives are available, is a good one. I suppose it often comes down to whether or not people realise the effects/annoyances/costs that different designs are potentially going to cause, in everyday life, over the life of the product. That was part of the point behind Norman’s book, I think – making these issues available to the public – but the fact that not even all designers think too deeply about it probably shows how big the problem is.

As for voting systems, perhaps someone could invent a wifi/bluetooth/GPRS sensor that fit into a watch/phone/PDA or dedicated pendant/bracelet? This would relay the skin temperature of the wearer such that it could be determined whether they were ‘hot’ or ‘cold’ and could adjust the heating accordingly. Naturally, those who have been in the heated area longer get more votes than those who’ve just arrived.

If distributed heat sources are too inefficient, then one could still use multiple thermostats and blowers to independently draw heat from a central source.

Rich is also correct that poorly-designed controllers are a problem. Problem is, even a well-designed thermostat will fail over the life of unit. A tax credit or government subsidy for annual thermostat calibration could help here.

In contrast to a normal house thermostat, where only a small number of people (who usually all know each other) experience the change and a consensus generally emerges, I was looking at how one might design a system for use in a space where lots of people might want different temperatures. Like the Rattle project I only had two relative options, warmer and cooler, although you might consider no action – ie, ‘I’m happy with the current temperature’ – as a choice too. If the preset was perfect, nobody would want it warmer or cooler; if it was only slightly off, only a few people (the pickiest ones) would request a change in temperature. If it was completely off, or had swung wildly one way or the other, even the most tolerant passengers would feel a need to make a request for hotter or colder. I also separated immediate feedback from the primary change the interaction would result in. I’d love to test the system and the underlying logic, but I’ve never had the opportunity.

Would be happy to hear any feedback on the project and the ideas in it – for what it’s worth Boeing and BA weren’t interested!

It’s quite common to get into a freezing car in the winter (or a hot one in the summer) and to want to get the temperature livable ASAP. In the car, cranking the temperature control to one extreme or the other WILL get the desired result.

There are two types of coolant-based heater controls in regular use: one is a coolant valve that controls the flow of coolant through the heater core. The other is an air-mixture valve that mixes cooler outside air with air blown across an always-on-full heater core. Both of these controls demonstrate the valve effect and (wrongly) train the human to think the home thermostat works like this.

On the “cool” side, the “Max A/C” button and air mixing controls reinforce the valve mental model…

DC

If that’s the way people spend money on a major discretionary expense, how much less can we expect heating controls to improve. Most are bought for the cheapest possible price by builders and never changed by householders unless they physically break down.