We've uploaded a short video showing the basics of how CityCAD works.
CityCAD is the leading conceptual design tool for professionals involved in large-scale masterplanning. It allows you to test hundreds of design options very quickly, and incorporates planning, sustainability, liveability and financial analysis. Best viewed in HDJuly 03, 2009
May 26, 2009
December 06, 2008
Getting Around London in 2030 ?
Here's a sketch diagram of what London's suburban rail network could have in 2030 :
(click for full size image)
This isn't based on any detailed feasibility study. It's the result of a bit of Saturday afternoon planner's megalomania - a loose collection of existing strategies and schemes, with a few changes. Most underground lines are left out for clarity.
Railways
London has lacked high frequency commuter lines which go deep into the city centre, like the Berlin S-Bahn and Paris RER.
The Crossrail lines will help remedy this, and make it easier to commute from a long distance directly to the centre without having to change at a main line station on to a congested tube.
There's also an opportunity to create an Outer Circle Line - following the route of the London Overground network - to relieve pressure on the centre, make it much easier to travel between the north and south suburbs and improve connections with the airports. This would be a similar scale to the JR Yamanote Line in Tokyo.
For an outer circle line to work, it needs good connections with the radial routes heading in to London.
There will be interchanges at Clapham Junction, West Hampstead, Dalston Junction, Whitechapel and Shadwell, but there could also be valuable interchanges at Willesden Junction with Crossrail and the main line, and at Surrey Canal Road with the Thameslink line.
A link at Willesden Junction could also allow high speed trains to run directly from Heathrow to either Europe or the Midlands and Scotland.
Airport Links
There is a proposal to build a new link between Staines and Heathrow Terminal 5, which could potentially be linked to Crossrail Line 2.
In the North East, there could potentially be a new interchange at Hackney or the lines could merge at another location, providing a direct rail link between Heathrow and Stansted Airports in under 2 hours.
A New Thames Gateway Airport
London is looking at the feasibility of building a new long-haul hub airport outside London, probably on an artificial island in the Thames Gateway.
I don't have enough information to know if it's the right idea or not.
It would improve the quality of life for hundreds of thousands of people in West London, it would allow a state-of-the-art airport with fewer delays and plenty of expansion room. It would also release a large well-connected site in West London for thousands of new houses (despite the current recession, new houses will still be needed in London).
On the other hand, it would mean tens of thousands of people who work at Heathrow would need to find new work or commute a long way to the new airport.
(When the main airport was replaced in other cities eg. Paris Orly to Charles de Gaulle, or New York La Guardia to JFK, the number of people affected was much lower, and Heathrow would represent a totally different scale of upheaval.)
There is also a huge cluster of services, hotels, logistics, car parks and other businesses in the area around Heathrow, which would also need to move. Lastly, the airport would be in a very inconvenient location. There might be a decent train service to London St Pancras, but getting there from the rest of the country would be difficult.
It would be expensive (My best guess is £25-30bn by the time it was finished), but in the long-term this could possibly be paid for by redevelopment of the existing Heathrow site and savings from fewer delays etc.
What do you think?
November 28, 2008
November 14, 2008
What is Parametric City Design ?
This year, at the Architecture Biennale in Venice, Patrik Schumacher gave a fascinating introduction to the concept of ‘parametricism’ as a new style, that could be applied to all design tasks including urbanism. As creators of parametric city design software, we feel this is a good time to produce a quick summary of what parametric design is, and to share our view about its possibilities and limitations in the field of city design.
What is parametric design?
Parametric design is an established concept in CAD.
Instead of drawing lines, a design is created by specifying parameters, and the actual geometry is then generated automatically by the software.
In the diagram below, the line on the left is just a line – it is drawn from one end point to another. The street designs to the right, however, were drawn parametrically. The positions of the end points were specified and the software then generated the pavements, vehicle lanes, central reservations and trees based on pre-set parameters.
The parameters in this case are the number, type and width of the elements of the streetscape - the pavements, cycle lanes, trees, vehicle lanes and central reservation, and their relationship to each other.
‘Parametric’ simply means ‘of or relating to a parameter’ - or in the case of CAD, generated by parameters.
What are the benefits of parametric design?
In architectural design and engineering, parametric modelling tools allow designers to create a framework of relationships and constraints between parts of a model. If one part is adjusted, then other parts will be updated automatically according to pre-defined rules.
In the example of the streets above, you could type in a new width for the pavement, and the model will be updated automatically.
The advantages are clear – things can be drawn faster, more accurately and with less chance of mistakes, and they can be modified extremely quickly. They can also be used to automatically extract data such as floor areas, or quantities of materials.
The disadvantages are that the designer is constrained by the rules of the CAD system, and changing the rules can require a high level of technical knowledge and programming skills.
The nature of city design and planning
All design exercises require a balance between creativity and analysis, between an intuitive approach and a systematic approach. City design and planning usually requires a much more systematic approach than other forms of design because of the quantity and complexity of the information involved.
It is often more of a political exercise and less of a technical challenge, as urban masterplans have to accommodate the complex, sometimes intangible, diverse and changing needs of a large number of people.
Also, because an urban masterplan can affect the quality of life of thousands of people, there is a clear obligation to choose a design approach with a proven track record and high probability of success.
So how can parametric design help city designers?
Parametric design has to be a systematic process because it is carried out by computer software - and in many ways it is ideally suited to city design.
Parametric design involves the automatic generation of design geometry, and a key question is what scale does this automatic generation take place?
At a relatively small scale, like the scale of a street or road, a designer can set parameters such as the width of the pavements or vehicle lanes, specify a path or alignment for the street, and the software can then help calculate the total area of paving stone needed, or the quantity of concrete needed, or the volumes of cut and fill to build the road through an existing terrain.
At the scale of a block, you could set parameters such as the number of storeys, the average unit size, and the depth of a perimeter block, and the software can then generate a visualization and calculate the total floor area of the block and the number of units within it. More advanced generative programs could optimize the height of a perimeter block so as to provide a specified number of hours of daylight to rooms on the ground floor of the block. (The number of hours of daylight would be the key parameter).
Parametric design can also be applied at a much larger scale – algorithms can be written and parameters set that will generate a large masterplan layout. For example, an algorithm could generate a street layout that varies the block size and density based on the distance to transport nodes.
What are the risks of using parametric techniques in city design?
The idea that we can generate an ideal city from a set of principles and logical steps has of course tempted city designers since long before computers were invented. However, generative parametric techniques make it easier than ever before to apply this approach.
It is easy to think that a generated/parametric design is legitimate just because it is based on an systematic process, especially if it happens to look beautiful on screen. However, as with all other software, the output is only as good as the program that creates it. Parametric techniques normally only deal with a narrow part of the city design brief and for this reason parametric technique should not, in themselves, confer legitimacy on a design.
Also, parametric techniques can leave little room for any kind of intuitive, emotional response to a site and its residents, and makes it easy to overlook or exclude vital characteristics that could make a design successful.
Having said this, there’s no reason why large scale generative parametric design shouldn’t be used to explore concepts, as long as it only informs the design process and does not drive it. Parametric software can be so powerful and produce such seductive images that it can require an effort of will (and common sense) from the designer in order to stay in control of it.
The difference between 'parametricism' and 'parametric city design'
The new proposed style of ‘parametricism’ is mainly an aesthetic agenda – assigning value to interarticulation and continuous surfaces while avoiding clear-cut zones/territories, repetition, straight lines, right angles and platonic or hermetic objects.
It’s true that 'parametricism' cannot exist without parametric software, but parametric software can, and does, exist without ‘parametricism’.
Parameters can be applied to rectangular city blocks and there is nothing inherently un-parametric about platonic solids or right angles. Clear-cut zones and territories sometimes work very well in urban planning, and can be dynamically generated by parameters as well. As Patrik Schumacher himself writes, designers "are employing parametric tools in ways which result in the maintenance of a modernist aesthetics, ie. using parametric modelling to inconspicuously absorb complexity."
Our approach to parametric city design software
In our city design software, geometry is only generated at a relatively small scale in the urban context – at the level of streets and blocks as opposed to whole urban areas.
In this way, the designer still relies on their experience and judgement to ‘draw’ the layout of streets and specify the size and shape of each building.
Once it is drawn, the parametric capabilities of the software can then analyse the model and automatically calculate floor areas, residential densities, car parking, energy and resource use, costs, values and other sustainability and quality of life indicators.
If a design is changed – if new storeys are added, or if a street is made wider, or if a land use is changed – then all of the floor areas and other data can be updated automatically, saving a huge amount of time.
When it comes to city design, the real value of parametric technology is not so much in generating geometry at a large scale, but in offering instantaneous feedback of masterplan design information and analysis during the design process.
June 05, 2008
Do City Designers need 'Le Correcteur Terminologique'?
I was sent this story from a month or two ago :
http://www.telegraph.co.uk/news/main.jhtml?xml=/news/2008/03/10/wfra310.xml
It's about a new piece of software that is being installed for the French civil service to try and discourage English phrases such as 'email' from creeping into the French language.
The word 'email' is understood by the whole world and is quite useful. Because of this, the new French software will probably make it more difficult for people to understand each other.
But it got us thinking about the problem of language in our own field of work.
All of the different groups involved in city design, planning and development speak their own dialects. For example, a 'house' might be called :
- a 'unit' by surveyors
- a 'dwellinghouse' by planners
- a 'dwelling' by architects
- a 'home' by sales brochures
A large residential development might be called :
- a 'masterplan' or 'place' by planners
- a 'new community' by a developer
- a 'new urban quarter' by sales brochures
- a 'dormitory estate' by opponents of the development.
A wide part of a street might be called :
- 'that bit of the road next to the new building' by locals
- a 'space' by architects
- the 'public realm' by planners
- an 'area of hard landscaping' by designers
- an 'off site infrastructure item' by a developer.
New shops in a town centre might be called :
- a 'shopping centre' by local residents
- a 'new piece of the city' by architects
- 'town centre retail-led regeneration' by developers
- an 'extension to the prime retail pitch' by surveyors
- a 'brownfield opportunity site' by planners
The fact that there are many different ways of describing exactly the same things illustrates how fragmented the city design, planning and development community is.
One professional group on its own is capable of creating a huge amount of impenetrable jargon.
When several professions get together, as they do in city design, the quantity of jargon can grow at a terrifying rate. Rob Cowan's Dictionary of Urbanism is a heroic attempt to control and explain this jargon.
Our urban design software, called CityCAD, was originally conceived as a way of trying to bring together all of the different groups involved in city design and planning.
It does this by creating a shared integrated model of the masterplan where all data can be stored. It doesn't solve the problem of language (it has to use some words, obviously) but the main technique for communicating information is through a 3D perspective view.
A 'Correcteur Termonologique' probably wouldn't help though, as everyone seems to enjoy jargon far too much to do away with it.
November 07, 2007
New 'Neighbourhoods' in Masterplans
What are Neighbourhood Units?
When designing New Towns or other large scale masterplans, a popular approach is to subdivide the masterplan into small 'neighbourhood units', sometimes called 'communities' or 'urban villages'.
Each Neighbourhood can be described as a circular blob with a diameter and an average density.
The diameter can be measured in metres, but it can be more useful to use 'Mins' Walk Across', as this is a more meaningful measure of distance to residents.
How many dwellings are in one Neighbourhood Unit?
This depends on its density and size. If the Neighbourhood is circular, this graph shows how many dwellings can be achieved at different densities and different sizes.
For example, if a new development needs to be 10 mins' walk across, then the graph shows that an average density of 30dph across the development will yield about 1000 dwellings, while an average of 40 will yield about 2000.
(This is obviously not an exact science, as the shape of the Neighbourhood Units will vary according to local constraints, and is unlikely to be exactly circular.)
NB. In the graph, the Densities are Gross (or 'Neighbourhood' or 'City') Densities. Normally it is Net Residential Density which is discussed and negotiated. Net Densities exclude major roads, hospitals, schools, playing fields etc in their calculation. Although areas of housing may have a Net density of 30 dw/ha, the overall Gross density of the Neighbourhood may only be 10dph or so.
What sort of services can each Neighbourhood Unit support?
This depends on a complex range of factors, but the number of dwellings is often a useful measure of what can be supported.
Let's say that each Neighbourhood Unit needs a service, such as a local shop, which needs approximately 2,000 dwellings to support it.
Going back to the graph, a horizontal line can be drawn at this number.
The graph tells us that a range of options are available to achieve 2,000 dwellings.
It could be 10 mins' walk across, but then an average density of 40dw/ha would be needed. This is a high Gross density, and probably only appropriate in a town centre location. Alternatively it could be 15 mins' walk across, with an average density of 20dw/ha.
Of course most developments are far more complex than this, but the graph is useful for quickly estimating numbers.
Our city modelling technology - CityCAD - will make analysis of complex masterplans with many different densities much easier at the conceptual stage.
September 11, 2007
In Defence of Bendy Buses...and more
Above: A modern replacement for the Routemaster? (see below)
In defence of bendy buses
According to today's Evening Standard, Boris Johnson vowed today that if he was elected Mayor of London, his first act would be to scrap bendy buses and replace them with a modern day Routemaster.
Bendy buses come in for a lot of criticism, and as a passenger on bendy buses for the last few years I feel moved to blog in their defence.
I think the greatest benefit of them is that they can deal with large numbers of people very quickly, and they do get you there faster.
The three sets of doors, and the ticketing, mean that people can get on and off extremely quickly. Anyone who doubts this should go down to London's Waterloo station at morning rush hour and watch the crowds streaming out of the station and on to the route 521 bendy bus. (The stream of people organizes itself into three queues, one for each set of doors at the bus stop. Buses arrive every few minutes, and the queues move forward much more quickly than they would otherwise).
I like the fact that bendy buses have plenty of room inside. There is room to walk around at non-peak times, and they almost feel more like a train than a bus.
Because of their size, they also tend to be driven more smoothly and much less aggressively than other buses, making the journey more comfortable and safe.
The approach to ticketing is controversial - you need to buy a ticket before you board, which is a hassle if you're in a rush and the ticket machine is broken. However, this is a policy issue which has nothing to do with the design of the bus. There are many other types of bus for which you need to buy a ticket in advance. And equally, if they wanted to they could easily introduce a conductor on the bendy bus who could check tickets and accept cash fares, just like the old Routemasters.
The question about cyclist safety is a serious one. Bendy buses (and any other long vehicles such as articulated lorries) obviously need to be sensible when overtaking cyclists, and shouldn't do it on a left-hand bend or corner.
In general though, all road users have to take responsibility for safety and I think the attitude of the drivers and riders is a much more important safety factor than the buses being slightly longer. It's worth pointing out that hundreds of cyclists act recklessly every day, while I've hardly ever seen a bus driving dangerously.
Yes, bendy buses have fewer seats and more standing room, but that's because it's an urban mass transit system - it's designed to move very large numbers of people relatively short distances, with lots of people getting on and off quickly.
So basically I don't mind bendy buses at all, and actually quite like them.
A new Routemaster?
Having said all that, the most interesting part of Boris Johnson's suggestion was the opportunity to create a replacement for the Routemaster - a new type of bus unique to London, which could once again become part of the city's identity.
We had a go at modernizing the form of the Routemaster (pls see image above) by making it slightly bigger and slightly more blobby and streamlined. Two points to make here:
- Tickets. The driver can't sell tickets to people getting on. So either people will need to buy a ticket before getting on (like the bendy buses), or there will need to be a conductor (like the old Routemasters).
- The Door. Many Londoners remember the adrenalin rush of running and leaping on to the rear platform just in time before the bus left, or jumping on and off in traffic. This was great fun, but extremely dangerous, and in the 21st century we felt we had to draw a sliding door at the back, which would open at bus stops only.
If London does build its own new bus type, should the new design try to emulate the shape of the Routemaster?
Perhaps not. The original was strikingly modern when first produced, and acquired its character over many decades as a backdrop to people's lives.
I feel there's a strong argument that the best vehicle designers in the country should produce something completely new and extraordinary - efficient, accessible, safe and fit for purpose but also unexpected, unique and distinctive.
July 23, 2007
8 Quick Tips for Urban Design Sketching
Introduction
Although at our company we spend most of our time creating computer modelling tools, freehand sketching is still an invaluable method of communicating ideas for urban design schemes.
Practice is obviously the most important thing when learning how to draw urban design sketches, but there are also many little 'tricks', techniques and 'things to know' which can help achieve more convincing drawings. We've summarized a few here:
Tip #1: Wobbly Lines
When drawing black and white sketches, it doesn't matter if the lines you draw aren't smooth - they can be 'wobbly' as the eye will tend to average out the line.
The most important thing is to get the start and end in the right place, and make sure than on average the line is heading in the right direction. If we concentrate too hard on making the line smooth and straight, it's easy for the line to end up in the wrong place.
When learning to do this, it can help to keep your hand completely off the paper and to draw the line fairly quickly. This can seem counter-intuitive at first (as we are used to writing carefully with the wrist on the paper) but it helps avoid rotation errors where the line becomes slightly curved.
Pls see the diagram below. It doesn't matter if the line is slightly wobbly, but it does look wrong if it ends up the wrong place. On the right, this describes a common problem - if the wrist is held on the page, then the hand will tend to rotate, and cause the line to curve.
Tip #2: Always Use Perspective Lines
Even for very quick sketches, it is always worth using vanishing points and perspective lines.
When starting out, it can be useful to lightly draw perspective lines with a pencil before going over it with a pen (and then rubbing out the pencil). They can easily add accuracy and credibility to a simple drawing.
Tip #3: Trust the Perspective Lines When Drawing
Sometimes the perspective lines don't feel instinctively right at first - in particular the angles of the tops of buildings may feel too steep. It's usually best to trust the perspective lines, and to follow them strictly until the drawing is finished. If it still looks wrong at the end, then try a different vanishing point.
Tip #4: Set the Vanishing Point at the Right Height
In most urban design sketches, the vanishing point and horizon are very low relative to the buildings (as people's eye level is low compared with most buildings).
Quite often, this doesn't feel quite right, and it's easy to accidentally draw the vanishing point a bit higher than it should be.
Also, when drawing places such as street cafes, remember that the horizon for people sitting down will be different from people standing up
Tip #5: Keep it Simple
Choose a relatively simple viewpoint to start with. Single-Point Perspectives (eg. looking straight down a street, or straight at a large building in a square) will save you a huge amount of hassle.
Don't worry about it looking too simple to begin with - it's easy to add in richness to the drawing later either by adding people/trees or adding detail to the buildings.
If the view you want is complex, or if it needs to be particularly accurate, then it may be best to create a 3D computer model, set up your view, print it out and trace over it. Google SketchUp™ is particularly useful for this.
If your view looks straight down a street, then our program Streetscape can be useful for getting the proportions right to start with.
Tip #6: Keep Practising Drawing Things From Life
Most people draw all the time in primary school, but then stop when becoming teenagers. However, our 'primary school drawing instinct' from those early years is very strong, and we are left with an set of mental images about how to draw certain things, such as the house below.
This primary school image of what buildings and other things look like is of course inaccurate, but lasts a long time and can subtely influence everything we draw. Drawing things from life helps to correct this instinct and replace it with an instinct for more realistic details and proportions.
Tip #7: Vehicles
Like human faces, modern cars are extremely complex 3D shapes and are very difficult to draw convincingly.
Because we are very familiar with the shapes, the slightest inaccuracy can become very obvious in a drawing. For quick sketches, drawing cars is best avoided unless absolutely necessary.
If you have to draw them...
- Drawing cars from above is easier than drawing them from the front, back or side. As a rule of thumb, they are three times as long as they are wide. The front headlights can usually just be seen from the top, and can indicate which direction the car is going in.
- Drawing from the front or back is easier than drawing them from the side. Distinguish between fronts and backs by the position of the number plate.
- If you have to draw them from the side or from any other angle, it's normally easiest to find a photo of the car you want and then trace over it.
Buses are easier - the only thing to watch out for is the proportions. We normally see buses from an oblique angle (at a bus stop), and so when drawing them from the side there is a natural tendency to draw them too short. Modern double-decker buses are usually 2-2.5 times as long as they are high, and they can look unusually long when viewed from the side at a distance.
Tip #8: Add Crowds
It's nearly always worth taking the time to draw a large number of people in your sketch, as it can bring an urban place to life even if the buildings are very simple. If you don't want to draw people close up you can normally get away with a wide-angle view such as this one, where people are little more than a scribble.
The key points to remember when drawing crowds are:
- Use a constant horizon to set the height of all the heads.
- Vary the size of the heads and height of people to create depth.
- Allow for a little variation in height (some people are taller than others) but not too much as the perspective may start to look wrong.
- Go to busy places and look at how crowds actually behave. Many people walk on their own, but there are also plenty of people in twos, threes and larger groups.
The tips in this blog post are short-cuts to achieving certain effects but as with anything, drawings always improve with practice. Please share any other tricks you may have come across...
June 28, 2007
Technology for Liveable Cities
I'm afraid we've been neglecting this blog a bit in the last month or two. There's a good reason for this though, as our team has been working hard on developing our new city design software.
We're interested in creating holistic integrated computer models of cities, which bring together all of the different specialisms - design, planning, financial, environmental, movement and liveability.
City Design on 'The Apprentice'
A couple of weeks ago, the last episode of The Apprentice was great fun to watch for anyone involved in the design and development industry.
The two finalists were given the task of coming up with a development proposal for a large site on London's South Bank. They had to work with architects and other advisors to come up with a scheme that was viable, but also high quality architecture.
The episode captured many of the issues that come up in large scale development projects, but we were particularly interested in how the teams dealt with the relationship between the design and the numbers.
On large masterplan projects, there is always a huge amount of information to deal with - for example floor areas, numbers of units and residential densities. The designer has a problem, as every time the design is changed, all of this data needs to be re-calculated. One of the key benefits of our software is that it will solve this problem and allow the design process to flow much more quickly.
The Dangers of Using Computers in City Design
We have several things in the pipeline for the rest of 2007. We're currently testing a new version of our street design tool called 'Streetscape Pro' and hope to make this available for free on our website very soon. Please contact us if you'd like to join the beta test programme.
We're currently investigating various ways that our CityCAD holistic modelling technology can help all kinds of masterplan projects - please sign up on our website if you'd like to be kept up to date.
Looking ahead, we're interested in exploring the positive and negative aspects of the use of computers in city design, and plan to organize an event later this year to explore these ideas in more detail.
We think that despite the obvious benefits of computer use, particularly in 3D work, there are some recurring problems:
- Misleading Accuracy - computers and print-outs can often suggest that a design is finished and resolved to the nearest millimetre when in fact it's only at an early stage.
- Limits to Design Geometry - like any medium, the processes and tools that people use to draw on computers can strongly influence the design itself, particularly if the design process is subject to time pressure.
- The 'Sketch-to-Digital' Interpretation - in the early stages of a project, designs will be produced by hand using pencil and paper. The process of translating these hand-drawn sketches into a digital drawing is a critical point, where the subtelty and elegance of shapes can sometimes be lost.
Please let us know if you have any views on this - we'll probably blog about the dangers of computers in more detail, together with some ways of avoiding trouble, later this month.
Recent Comments