Rather than try and do a massive re-edit of my script for the presentation I gave at the GeoDesign Summit this year I thought I’d just cut and paste it straight in. Yes I didn’t say it all exactly like this and somethings were added as I spoke, but you’ll have to wait for the video from ESRI for that. Until then here is the raw un-edited script from my presentation.The slides look odd because in the presentation they were animated, here they are exported as jpegs at the end state… I had thought about exporting as video and reading over the top. We’ll see if I have time…WARNING! This took me 20-25 minutes to read so go make yourself a tea or coffee before reading…
Thank you for inviting me here. It’s exciting to see the presentations as well as to meet some of you. This summit already has my head buzzing with new ideas and ways to improve how I work. So far it looks like I’ve got a lot of work to do when I get home.
My name is Elliot and I’m a Director of Garsdale Design Limited. We’re a small family run professional planning and architectural consultancy that specialises in Middle East city master planning. As you can see we’re not exactly a large multi-national. Our office seen here is a converted traditional Yorkshire dales barn.
If we start with an outline of my presentation today.
So in this presentation I am going to stick to talking about the workflows and practices I know best, rather than talk to you about the academic theory behind them.
As I work in a small consultancy, we’re concerned with a lot of practical things. For instance managing our workloads and juggling our multiple roles, as inevitably characterises a small office.
On a professional level workflows are what really concern me. I’m interested in how we can extract value from our digital work and get from A to B.
This is really a presentation about my journey and how relatively recently my whole perspective on how I work has been turned on its head.
So please follow me through the journey I’ve taken.
There are two types of workflow that take place in our office. Firstly a main project workflow which is linear with defined start and end points, often designated by the client and the project’s terms of reference.
Secondly there are those process workflows that are often the small iterative evaluations associated with particular tasks. These might be population modelling in spreadsheets, or at the detailed design stages, they include sketching and CAD work.
We generally finish one phase of a project and use the outputs from that to inform the next phase.
Ultimately the smooth flow of data through our systems are critical. It’s often when different types of work or data need to come together that it gets difficult.
So let me start with the main work we do now and the workflows that we have been using.
In my office we undertake a variety of work from the very small single residential development up to region wide structure plans.
But the projects that define and consume us at the moment, are city master planning projects. Currently these are in Iraq.
As you can see from this slide we have a fairly linear approach to the process. This is not only defined by the client and the contract, but is also how we would traditionally do many planning projects. Each stage in this example has a proper report with associated spreadsheet tables, writing and of course mapping.
What many would see as the creative work is in the later phases where we start designing urban areas, either at a strategic or detailed level. Sorry to all you spreadsheet power users out there, I know you can get creative too!
Our traditional workflows do work, but increasingly they are buckling under the weight of information we now have at our finger tips, thanks to the power of technology and of course, GIS.
Sometimes the amount of incoming data for a project feels a bit like this… At the beginning of a project there maybe drought, but that quickly gives way to a torrent.
ArcGIS is used to manage all of our geographic data that comes from a variety of sources. Householder surveys, old AutoCAD drawings and often quite a lot of shapefiles that have been digitised off old paper maps.
Difficulties such as data quality are inevitable with our diverse projects. Our Iraqi partners do amazing work translating Arabic data into English, but we still have some guess work on our hands to extract the value of their work.
Managing the changing nature of some of our data is one of key challenges we face.
Needless to say cities don’t stand still, politely waiting for our project to finish.
Especially in Iraq where there has been a desperate need for infrastructure projects. It’s a brave person who stands up and says “don’t build that sewage treatment plant or road until the master plan is ready, oh and that will take a couple of years.”
So our work goes in tandem with the normal growth of a city, as well as major construction and infrastructure projects. This guarantees that our master plan is out of date before it is adopted. But don’t worry, part of these projects is about training the Iraqi planning departments to manage our master plans knowing how to update them. In a real sense we teach them to think in terms of ‘process’ rather than ‘product’. A master plan is not like a ‘blueprint’!
Our clients in Iraq are mostly planners and politicians. Like in any country they are educated, active and eager to improve their cities, albeit in difficult sometimes hostile environments. Local planning committee meetings can be caught fraught affairs!
Client requests are part of the process. As the project progresses and plans become more evolved, the planning teams in the cities involved send in requests for advice on certain sites to ensure compliance with the emerging master plan.
The photos that are shown here are a recent visit of our clients and Iraqi partners who came to the UK for training, (you can see our team is a family affair as you can see our youngest colleagues there, I mean my two daughters!)
All our projects of course, have deadlines. But the nature of development means that our approach has evolved to be flexible. We always expect changes and inevitably many occur just prior to a submission.
Throughout our projects we have always looked at ways to speed up our visual presentations. In terms of workload, they can take up quite a bit of time but are relatively low in importance when it comes to the actual analytical side of our work.
3D imagery has always been part of the visual mix and clients get excited by it and have now come to expect such visualisations. Try handing in just a 2D map to a client and they are guaranteed to ask how this would actually look.
We recently completed the Nasiriyah City Master Plan in Dhi Qar Province in the south of Iraq and are currently in the middle of preparing master plans for three other cities in Iraq. In fact I was working on them prior to coming here.
This means that we need to have a wide variety of visualisations produced. Towards the end of these projects we produce a range of presentational materials to be included in the final report including wall display maps and 3D visualisations.
Our final city master plans also include a set of detailed studies in the final phase. These can be new neighbourhood plans, linear development plans along major transport routes or, for example, detailed urban renewal schemes for waterfronts.
Most of our visualisation work has a relatively simple, but time consuming workflow. Increasingly clients want sophisticated 3D visualisations in a variety of media, from the printed report, to websites, video, and fully interactive walkthroughs.
They do not just want to be told in words how to implement it – they want to see how it will look when implemented and this is perfectly understandable. After all most plans are subject to consultation by the general public too.
So how have we done such visualisations?
- Scan and georeference,
- import into sketchup,
- import into rendering software.
The different methods and workflows of producing these visualisations add to the time scales just at the point when we need them to be completed really quickly.
We have always used a combination of professional software packages to produce our 3D visualisations. We have always used traditional hand-drawn images, but then increasingly SketchUp and for the presentational material, but these can take considerable time to complete in-house.
What we have never done before is produce a complete 3D model of a whole neighbourhood, let alone a city, because such a task seemed to be just too time consuming. Conventional professional 3D modelling packages are good and can do this, but they require too much of our time and effort for anything more than visualising small sections or details of our master plans.
After some research a couple of years back to tackle such issues I discovered a useful urban modelling software tool called CityEngine. It showed promise purely because it worked with a variety of formats, so I thought it could fit our AutoCAD/ArcGIS workflow nicely.
In the beginning we used real data from a project and essentially experimented with it. But from the moment I started using it, it became apparent that this ‘Procedural Modeller’ would not only save us huge amounts of time producing 3D visualisations. It had the potential to totally change the way we think about our workflows.
Well, this is exciting as I am finally using CityEngine in the way I want. We’re combining 3D work in SketchUp with some of the models in CityEngine. As you can see the time taken to model urban areas has dramatically increased.
- 4 days in SketchUp for a neighbourhood centre
½ a day in CityEngine for a whole city quarter
It really has been that dramatic, of course we’re now looking to combine our detailed models in SketchUp from previous projects with CityEngine.
The other tool that has been helping us is a real time rendering package. We are using Lumion at present. While secondary to the modelling Lumion improves the quality of imagery we produce and we’re no longer waiting overnight to see the results.
Everyone has worked on a project where changes are needed just prior to finishing it. This slide shows what can happen on a project, they build a bridge, without telling us! We might have suggested it be better located elsewhere… but it’s too late now.
We’re discovering at the detailed design stages that CityEngine can allow us to make quick changes to the detailed plans, such as a new road, and give us a corresponding 3D model. This sort of change would have taken significant time a few years ago but now we can do it in less than a day.
I’ve got ideas for CityEngine. Our next job is an urban renewal project which willl require much more detail than the relatively simple level of detail we have used in our master planning work. CItyEngine seems for us, at any rate, to be able to model at the macro and micro planning levels.
But while I’m not involved in its development I’m sure there are more capabilities going to be developed for this tool. I’ve been working with ways of letting the rule file create it’s own data instead of respond to existing. By that I mean choosing points within a city and varying the model depending on its distance from that point. The beauty here is I can have multiple points. Could I really use CityEngine to distribute Neighbourhood Centres based on variables I give it? We’ll have to see.
But until then try not to think of CityEngine as being just about cities, it has much more potential. I’ve already looked at how it might help respond to soils and relief in a forestry model so why not other areas as well?
So finally I come to the term GeoDesign. I’ve been wary of using the term in this presentation. For one thing it didn’t feature on my university courses, so I don’t know at first hand how it is regarded academically. The other reason is I believe a lot of us have been doing ‘geodesign’ for sometime, but never explicitly using the term ‘geodesign’ to describe it.
When I was invited to be a featured speaker I thought I better read up on the GeoDesign term and one of the first things I read was linked to from this summits website “Changing Geography by Design”, one particular phrase that stood out for me was “The Sketch Concept is key to GeoDesign”.
In some respects I should worry! If the key to geodesign is sketching I’m in trouble. You see I can’t draw! No really, I’m quite good at tracing and give me a colouring book I’m reasonable. But sketching and drawing, well I’m not an artist, my 3 year old daughter is!
Both of my fellow directors, can draw but they’re architects by training. Not only that but they have at least 30 plus years of design experience as well as being used to working without computers. (yes there is some still around…)
So is it back to the old drawing board?
For all my colleagues artistic talent, today’s projects use only a fraction of that. We’re increasingly reliant on vast amounts of facts and figures, as well new theories. The sheer amount of data we now have to take into account can overwhelm the design stages. So the design process has been forced from pen and paper on to the computer. For my colleagues generation design on computer is not natural. So what happens is this: They draw designs and I scan them, georeference and trace their designs into a GIS.
It’s lucky I’m good at tracing!
As I see it the future of Geodesign and the emergence of devices like tablets will allow people who are good at sketching to get back to the pure design. Designers will then know that what they do is constantly being informed by underlying data, and that if a key assumption changes, their designs can still be used within a new model.
I suspect I shouldn’t be worried at my lack of design expertise: maybe a geodesign program could well be created that will make it look like I can draw! Until people create that ‘geodesigner’ my strength is in systems, workflows and technologies: theirs is in their ability to conceptualize and ultimately produce a good design.
So my hope is that in the near future we’ll be able to a have a natural, almost effortless approach to design. This should look as natural as my daughter is doing here, but actually we’re harnessing huge amounts of data and knowledge as we do so. In fact just like we already do but, on a bigger and better scale.
So this is my ‘big idea’ because our workflows are going to change radically in the next few years. Look at what CityEngine has done for one of our workflows. It has already cut the time spent creating a model by more than half and allowed us to keep our 3D modelling in house. It will only get quicker, as our stock of rule files increases and the software development moves on.
For me the use of CityEngine, the ideas of Procedural Modelling and the concept of GeoDesign have now come together. Let me try and explain….
It started with me wondering what a really exciting demonstration of our skills and the power of CityEngine would look like in the future?
Such as demonstration would show how the use of a tool like CityEngine can inform at all stages of a master planning process even from the outset. This sounds strange but at present we start with background studies and evolve that into a number of growth options for a city and from that the client chooses. This happens over months if not years!
The problem with this approach is that early on the client chooses their direction of growth, but what if the situation changes? It’s going to be costly and timely to start again.
What I envisage happening in the future is that tools like CityEngine and the concepts of GeoDesign can be harnessed together. This will allow multiple versions of a city master plan to go through all at the same time. Then the client could choose at the end of the process a complete city plan, knowing all its implications from Building Information Modelling (BIM) to any other measurable indicator.
Essentially you could start with a blank sheet knowing how many people you need to plan for and a default set of planning standards based on the country or region the project is in.
As the project progresses and more information is delivered, existing roads, terrain, or landuses for example, and as each data layer is imported into the GIS, the CityEngine rules you have created run over it again and within your defined parameters remake the proposed city/growth area. As our ideas are formulated, new planning policies can be created that are fed into the rule files so that a very specific model for the area is created to show what the resulting urban fabric would look like.
For example you could take solar data and use that as the basis for orientating buildings in the master plan. So with a clever rule whole buildings could be orientated automatically depending on their location in relation to the sun, even in a valley! I already do this for the visualizations with details such as satellite dishes and mosques that need the correct orientation on a city wide level.
The trouble is that the concept is simple but it requires lots of computing power. However I think we’re approaching now a time when computing power is a product of how many machines in the cloud you can afford to have working on a problem. The Instant City idea relies on cloud computing (or a very powerful computer) to make it a reality.
What I envisage is a large visual display at a “Geo” conference and many tablets available. Attendees could contribute to the design process of a city by adding data (satellite imagery, obstacle maps, roads etc..). As each layer was placed in the model, CityEngine would update the city model and GIS would in turn update its analysis (pie graphs etc…) all in real-time.
This is very much a formative idea at the moment, and would need a lot of development. Nevertheless it would be a great experiment and show off in real time some of the GeoDesign concepts in practice.
So I have some thoughts about the future.
- Access to “OpenData” will impact on us all
- GIS Software is getting easier & cheaper
- Tying GeoDesign into “haptic” feedback
- “Gamification” ideas (Thanks to a recent James Fee blog post for the Minecraft reference)
- Large rollable displays
- Fibre Optic to the premises for all is now essential
When we start managing large amounts of data using future Geodesign tools suddenly the PC, as a platform is inadequate.
We will buy our processing time from the ‘cloud’. Time becomes only a question of how much computing power are you willing to purchase.
However internet connections where we are and I expect for many of you is inadequate particularly in the upload speeds.
So in my free time I am involved in a community project to bring fibre optic internet connections to some of the more remote parts of our area in the Yorkshire Dales.
Are my ideas about the Instant City impossible to realise?
Well, if you had told me that someone could build a 30 storey skyscraper in 15 days, a month ago I might well have given you an odd look I probably would have said “Almost anything is possible but to pull it off would require a lot of effort and some interesting thinking.”
I think the people who built a 30 storey skyscraper in 15 days might have something to say about that. Oh and look they are planning a 220 storey skyscraper in 90 days….
Before I finish I thought you might like some video of the CityEngine work I’ve been doing. It’s speeded up obviously but what I’m doing here is changing us. This is all work done in one afternoon. I am trying out the look for different rule files on an entire city quarter.
- First electricity lines with pylons.
- Next a generic office block.
- Satellite dish orientation
- Tram anyone?
- What about a parking garage? Cars or not?
- Adjusting details of one house type.
- Different road classes and associated paving/onstreet parking?
- How about a new set of house types for the whole model?
- My last ones is a favourite, here’s that bridge being made, depending on the rule file I can give it square or round columns and more…
For those who like some facts and figures:
- The model is on approximately 200 hectares
- includes about 1100 homes of different types
- 877 street trees of various varieties
- 5832 date palms within the gardens of those homes
- 877 natural, in other words randomly scattered palm trees in unused space.
- 64 Tram pylons.
- 818 street signs.
- 934 watertanks on the roofs of those houses
- as well as 1736 satellite dishes.