We’re moving in a couple of weeks to a house with oil heat – which has me worried about heating costs. I’d like to be able to manage the heating of the house via a climate server that has internet access.
The mimimul functionality this server will need is temperature readings from various parts of the house and control over the furnace valves that control where the heated water goes that heats these areas.
The next level of complexity comes with a weather monitoring station with data available to the server (like this one) that can provide exterior temperature information. I would like to use this to predict how much it will cost to keep the house at a specific temperature on a specific day.
The server will surely be online – with access to forecast model predictions of expected temperatures. I would like to use this in the modeling of the household environment and predicting heating costs.
Another layer of complexity comes in with the central air conditioning system – which has two compressors outside the house. Balancing these two units such that the optimal cooling is achieved would be something I would like to monitor and control on the server.
Eventually I like the whole thing to be driven not by a clock (morning vs. night), or even a weekly schedule (weekend vs. weekday) – but to be tied directly to the family calendar, such that heating and cooling needs are specifically spelled out – and a house based use case like:
- M@ is working at home today
- SuzShe is working the morning and coming home with the Boy at 2pm
- The Girl gets off the bus at 3pm
- Red Sox game is at 7:05pm
Would be stacked up against a hypothetical weather forecast
- 55deg at 8am
- 75deg at 12pm
- 65deg at 5pm
- Mostly cloudy in the morning
- Overcast in the afternoon
Would be balanced with the
- current temperature
- the relative “cost” of heating/cooling various rooms in the house
And could potentially not just control the heating and cooling systems with control outputs like
- 6am-9am upstairs at living temperature
- 7am-6pm office at living temperature
- 2pm-6pm kitchen at living temperature
- 3pm-6pm playroom at living temperature
- 7pm-10pm media room at living temperature
- 7pm beer storage at beer serving temperature
- 7:30pm upstairs at sleeping temperature
- 8pm red wine storage at bordeaux serving temperature
And the data would be presented in such a way we could see things like
- M@ should work in the basement until noon, saving $8 in climate control
- The children should do their homework in the media room, saving $3 in climate control
- Dinner should be cooked on the grill and eaten outside on the West edge of the deck, saving $1 in climate control and with a 82% chance of seeing a good sunset
And the really advanced piece of programming would spit out things like:
- Today’s climate control cost is estimated at $14.
- Working from the office is estimated to cost an additional $4.
- The pool party you declined to goto at Vader’s house would have saved you $8.20 in climate control
I’d also like the system to warehouse data, so I can perform query’s that might tell me that running an electric space heater in my new office will be more efficent than burning oil. I will definatly be interesteed in how much a pellete burning stove could save on our heating – but only after knowing how our daily heating needs fluctuate. What I’d really like to produce is data supporting a geo-thermal heating system.
Pie-in-the-sky uses include having our family menu detail if we are sitting in the dining room, the kitchen island or (gasp) eating in front of the TV and controlling the interior climate appropriatly.
And now I must pack the garage.