A conversation about data with Phil Jones, external consultant for the Pushing Doing group, Hannah Jones, 17th February 2007

Phil identified six possible types of data that can be collected to determine whether or not a building will work. These are

Analytical data which is produced by the laws of science - if you know the properties of a material and you know the temperature either side then you can determine how much heat passes through it. But this can only determine fairly simple systems, not as complex as buildings.

Empirical data, taking measurements - you take the temperatures in a building but that has error because measurements always have error and you have to create the boundary conditions for it.

Experiential data is data gathered from experiencing doing what you do for complex systems anyway - people tend to look for agreement between people and models and analytical data but you cannot say that any of it is the truth it is just that this is what everybody agrees that is their perceived truth but because people base it on the same sort of assumptions then it might not be the truth.

Data from complex modelling might exist in between empirical and experiential data.

Maybe after experiential data comes Intuitive data which is looking at unknowns.

maybe Imbedded~~ or ~~green:Implicated data is another type of data almost the opposite of experiential data

What Data do you Need?

When you are designing the energy conditions for a building you start with a conceptual model and you decide the variables and the parameters that you put into it. Then you ask 'what data do you need?', and 'What is the criteria for success?' Then from the conceptual model we usually develop numerical models and so everything is done numerically on computers, lots of calculations done very quickly rather than fitting something to an equation. So it is done with simple calculations but millions and millions of them. For example, you calculate the exchange of heat between one layer of wall and another every couple of seconds for a year and you do that for every wall in the building and then for the other components in the building and that is the way you usually develop models.

Lets say we are interested in seeing what the energy use is in a building for a year of being heated. Everybody could look at their models and they could say how close are we to each other. You can’t do that analytically because in complex systems there are always assumptions and behavioural factors and combinations of different analytical solutions that might be configured in different ways but you can compare it to empirical data but then that might be wrong because measurements are always wrong and they are only for a certain set of conditions.

The Inter-relationship between Data and Information

Information can become data, a given for another process. For example you might have a building scientist collecting data on the properties of materials and the temperature to determine the effectiveness of the building. Then you might have a building developer who is using this data on energy use to make a decision about whether to do something about the building to make it more efficient given other information about costs and other things, so that energy use is data for another process. Data to information can be a very complex idea when you look at decision making, it goes from one to another. As a building scientist you are a data collector, a data user to produce information to aid a decision, and then you evaluate this process and determine whether or not you need more data to make a more informed decision.

Knowledge Drift

Some people say you can offset the depletion of resource capital with the increase of human capital/ social capital because you become ‘clever’. But there is a danger of knowledge drift where you get all this experiential knowledge building up and it drifts away from whatever the truth is and all of a sudden everyone is doing something wrong and it has to come back, there is a step change. What happens in design a lot is that people don’t do enough of the analytical and empirical data collection and the complex modelling. They base most things on the experiential data/ information and that tends to drift in the direction that people want it to go. Very often designers will say ‘Oh I’ve seen one of those and I want it to work on my building’ so they assume it will work on their building because they want it for aesthetic reasons or they want it to make a statement and so your manipulating information and if enough people do that then the knowledge drifts away from the truth and then after a while something will go wrong and you have to pull it back into some stream. In the pushing and doing the pushing is about knowledge being driven from various forms of information like analytical, empirical, complex modelling and experiential but then there could be a pull into a particular direction. For example, if George Bush doesn’t want climate change to be a reality he creates another set of paradigms that he then gathers information for, its like tobacco companies saying smoking doesn’t do any harm, they create this myth and they create a knowledge base to support the myth and so it becomes a knowledge pull in a particular direction to suit a political or economic gain. With designers they have an idea and they want it to work and so they manipulate information to make it work.

Synergy is an Elegant Solution

Synergy is an elegant solution to a particular problem. Some people design low energy buildings that work but they look like a camel.

If you want to produce four billion possible solutions to the design of a building, consider the possibility that if you make certain decisions they will answer 99.9 of those four billion and they will all be taken care of in a synergistic way by default, if you look at the way buildings have evolved that is what has happened. If you look at traditional, vernacular buildings the problems have been eradicated by evolution and so what you get is a solution that people don’t need to think about, it happens because they do it in a particular way. The problem with building today is that we prototype everything and we don’t build on experience and so we are trying to do a task that is far too complicated to do at once and what we have to do is to learn from the vernacular approach but applying modern materials or uses. Maybe you could call that imbedded dta/ information, information that is imbedded in solutions that are hidden almost or not thought about at the time so a lot of the data, a lot of the information is imbedded in a solution that you use. The information is has been derived in an evolutionary way so you don’t need to generate the data every time you need the information. You could say if I use a lot of thermal mass in a building then that has a number of consequences, structurally it soaks up the heat gains so it provides stability, it creates an image of a building that has an aesthetic, but you don’t think of all those things, you think heavy weight building and all these other things fall into line without even thinking. You might have to fine tune them a bit. Its like saying this building is going to be naturally ventilated, if you say I’m going to have a sloping roof that has many implications (implicated data?) but all that information is imbedded in the sloping roof, its like a pattern recognition and pattern recognition is all about imbedded information. So maybe imbedded data/ information is another type almost the opposite of experiential.

return to Pushing Doing
return to Main m21 Page
return / go to Group Coordinators
return / go to Contact Details
return / go to m21 Events Calendar