Tip:
Highlight text to annotate it
X
So they input two or three things; output and APC,
but behind it- real power.
Here's another example where people might be interested
in the domestic market of knowing what control scenario to pursue.
So it's quite possible to configure models of the domestic sector,
lots of them,
and then to pre-simulate all the different
control permutations; keep all those results away,
and then let them come in and say "What would
happen if I had this type of control?
Upgrade my house to that."
And you go off and you give a dynamic result without
any need to define the size of the radiator,
the length of the pipes and so on...
Similarly here,
if you want to design a biomass boiler; all the
permutations have been done for that - and a recent one
we did in CIC Start Online was a "post - occupancy evaluation"
where you go in and you find out there's a problem and
then comes the rub - "what do we do?"
So you can have pre-simulated actions,
so you're really just going in and saying
"well I think I'll pick that one."
And you've already quantified it out.
So they're all dynamic models,
but you never need to feed the data required.
To just finish the modelling side,
what you're going to see more and more in future is
in our homes and offices and so on,
with internet access,
we are going to buy services.
At the moment you download a video - it's a service.
Care for the elderly,
where you look at temperature and movement,
to protect people in buildings,
more security.
But you can start to have all sorts of energy services
- environmental monitoring and smart metering,
and local and aggregate control where a signal goes out
"all fridges in the east-end of Glasgow,
off for 45 seconds - as long as the compartment
temperature is below something..."
- and that's equivalent to not starting a
turbine at some local power station.
And if you allow that in your home,
you get a better preferential tariff.
And these will be delivered using internet or
wireless infrastructure.
But they'll be delivered by service companies
using these modelling
tools to know what's the best action to take
before sending out the actuation signals.
In order to achieve this world,
we need a modelling based design evolution.
This is a complicated slide,
it's a very simple method.
We can't just put programmes into offices
- we need someone to control these methods.
I call it a "performance assessment method".
RED is action,
YELLOW is knowledge.
So we establish a model for an unconstrained base case.
So someone in the office says "Where are we going to start?"
Then we've got to calibrate the model using a reliable technique
- "Who's responsible for calibrating it?; what's the reliable technique?"
Assign boundary conditions of appropriate severity
- what are we trying to do here?
So these are methodological questions and these can be set up
for everything we might conceivably want to do.
So my question is "How might we deliver that into practice?"
And my answer is "embed all of this into CAD” - so that as you have a
session with CAD there will come critical trigger points
where the little postage stamps on my previous slide can be fired off.
The fact that there's a computation agent doing a complete
3 dimensional multichromatic lighting simulation
- you didn't ask for it,
all you did was provide geometry and construction;
and it knows it's got sufficient information to spawn that off.
So a little while later,
a line will appear on the screen "this is what it will look like"
or "this is the daylight you could get access to".
So it's a different model to the one we've known in the past.
And I think this could be well embedded within CAD.
To embed high fidelity tools (which they are) within work practices
in a manner that supports virtual design through the
interactive manipulation of the design hypothesis with the
performance feedback in real time.
You're just talking - suddenly you're getting performance feedback.
It's real time in the sense that it's timely.
It's not like you say "Do a lighting simulation"
and you get the answer NOW,
but you're working away on something and then up pops
the results of a lighting simulation.
And these would be the adaptations that would be required to the CAD bim.
Because historically CAD is not concerned with integrated
building performance simulation.
Integrated building performance simulation is the superset dataset.
CAD is a very weak subset.
But - a little speed read of that,
all the different things we would need to add in because that
is what integrated building simulation requires.
And just three little technical things to finish.
First thing we need to do is change the data model of CAD.
IBPS is a superset, CAD is the best agent,
however, for manipulating the data model,
and this is a NYAM representation of data and they have been
well researched these representations.
We know what they are and all we need to do is inside the
CAD system extend the data model and lots of people are working on that now.
They tend to be working in the areas which are not the quite
ones on the previous slide,
but it is not a rocket science to move it in that direction.
Second thing is, we need to change the user interface.
We need to separate out the various aspects of how we talk to users,
handle the dialogue,
because at some stage in CAD thing you are going to say
“I need to have a little discussion now about the control systems”.
And you can’t suddenly do that in the middle of an
architectural dialogue where the architect say
“I do not anything about this”.
So, there is got to be a recognition that we
separate out the way we have dialogues,
where we place the knowledge,
how to handle users,
how to handle applications.
The other area that would be required in CAD would be an
explicit process model which defines formally the workflow
and performance assessment approaches to be undertaken.
What applications should be used,
by whom and when,
and against what performance assessment methodology as previously explained?
And again,
Cad is the best agent for this and there is been significant
work undertaken to date to look at the best way of representing
the process and making the process model explicit.
And, these types of formalisms could be readily
embedded within CAD to control that process.
So, where to for the practitioner or the student who wishes
to start the long voyage into this world of explicit
simulation emulating future reality at the design stage?
I’ll bring to your attention a couple of possibilities.
One is the existence of the International Building Performance
Simulation Association,
and this organisation has existed now for some time and has
regional affiliates throughout the world with new regional
affiliates being set up each year.
So, it might be possible to contact affiliate in your region and
become engaged with likeminded individuals trying to see where
and how this powerful new design tool might be applied.
There are other initiatives,
particularly here in Scotland with
European Regional Development Funding such as CIC Start Online,
such as the Scottish Energy Systems Group,
such as RenewNet which is concerned with renewables.
And, these organisations can provide the kind of initial funding
which could be used for a practice to get some experience i
n explicit computational modeling.
So, to conclude my presentation today,
the real need,
as long ago recognized by Robert Shaw,
is to leave the future generations with knowledge and capital
such as that they can obtain quality of life at least as good as ours,
all in all.
So, against that definition of sustainability I would suggest
that the best way forward is to pay attention firstly to
environmental quality and energy efficiency.
That is the prerequisite of all that follows.
To take steps to make sure that we can prolong the useful life
of the fossil fuels, of course,
making them cleaner in the process.
And, those two things together can buy us the necessary time to
bring forward the new and renewable energy systems solutions
which will take some time do deploy at the scales required.
In order to make good judgments,
effective judgments,
take effective action,
that would require much more powerful tools than the profession
has previously used and I believe that those tools should be
integrated building performance simulation embedded within CAD,
CAD being therefore the agent of control.
So that brings me to the end of my presentation today and I would l
ike to thank you all very much for listening in and I hope you each
take something of value from what I have discussed today.
Thank you.