3 ms·
Agreed. I'm a masters of architecture candidate working in this area[1] and I would add that to engage in contemporary urban theory would require a much more so
by saeranv 12y ago
Agreed. I'm a masters of architecture candidate working in this area[1] and I would add that to engage in contemporary urban theory would require a much more sophisticated urban simulator than is achievable with the current state-of-the-art. Contemporary urban theory goes beyond issues like outdoor comfort and appropriate formal scale (which is what this blog post seems concerned about) to additional issues like hyper and mega-city growth and resource conservation. Taking this into account requires multiple simulations that would be able to model the inherent trade-offs of urban design.
Traditionally there has been software like Esri's CityEngine - http://www.esri.com/software/cityengine http://www.esri.com/software/cityengine - which procedurally generates[2] urban geometries based on user-defined typologies and land-use. However CityEngine is just a geometry generator, and therefore far too basic to be used for contemporary design which must also negotiate factors like building daylight accessibility or positioning transit nodes for optimum walkability.
A decent urban simulator therefore needs not only a procedural city geometry generator, but also simulators for walkability, building operational energy use and environmental simulators (ventilation, solar radiation etc). MIT's Sustainable Design Lab is developing the first (I believe) serious integrated urban modelling platform: Urban Modelling Interface (UMI) - http://urbanmodellinginterface.ning.com/ http://urbanmodellinginterface.ning.com/. Unfortunately UMI doesn't at this point include a way to procedurally generate urban geometries.
A combination of CityEngine (with a much friendlier UI) and UMI would provide the kind of complex simulation environment that I think a realistic urban simulation game would require.
[1] https://github.com/saeranv/cactus https://github.com/saeranv/cactus (Almost have a proof-of-concept ready for this).
[2] Using a shape-grammar engine - http://en.wikipedia.org/wiki/Shape_grammar http://en.wikipedia.org/wiki/Shape_grammar - which operates in a recursive manner and is thus a reasonable imitation of actual urban growth.
- mortenjorck 12y agoI would eat a game like that up. SimCity has so wholly established the popular concept of what a city simulator should be that everything else, from Cities XL to Cities in Motion, is inevitably built on that template. Don't get me wrong, the SimCity series is great, but the domain of urban simulation could be vastly wider than what fits within the boundaries of the Maxis tradition. Different things could be abstracted; other things could be simulated in greater depth. I'd love to see, for instance, less reliance on detailed buildings and more reliance on realistic building plans. Or less direct control over construction and zoning, having to go through councils and referenda, or backroom deals to get things done (imagine all these things having various stats for rolls like an RPG).