Populating the Digital Earth - Its Expensive

Tweet

Six years ago we carried out a review of digital cities around the world for the City of London. Their question was how much do digital cities cost and should London have one.

The cost of building digital cities is considerable, one architects practice has spent approximately £10 million for a 3D model of the city, ours cost slightly less (see our Virtual London series of posts for more details). Cost is relative to accuracy and detail, yet with technology moving on at an ever increasing rate algorithms are becoming available that interpret LiDAR and oblique aerial photography to semi-automatically create cities.

The movie below illustrates raw 1 metre resolution LiDAR data for London in 3D Studio Max:



Google Earth left us wide mouthed when it launched its 3D cities and its community led approach of using SketchUp and the 3D warehouse tackles the cost to some extent by using the user base to build the city. This ties in with the whole idea of Wikinomics, recently the subject of an excellent book by Don Tapscott, indeed SketchUp is, in our view, one of the most important pieces of software in digital modelling to ever be released.

The movie below illustrates some of our work for Newham Council, London - note the move to photorealism half way through:



Microsoft however has taken the money route - that's not to say Google Earth is low cost- £400 million a year to maintain the data alone, according to a recent conference speech by ESRI. In a comment posted on their development blog with regards our series of posts and Franks Taylors thoughts on the Google Earth Blog they state that in relation to the building of Virtual Earth's 3D cities...

this is a very expensive process to get off the ground. The startup costs are astronomical and the complexity is daunting, but it scales very well - the cost per building comes way down if you plan to create models of 1000's of buildings (and that's just one city!). If you only wanted to model 1 or a handful of buildings you would grab your favorite 3d modeling package and have at it. But if you want to model all of say Manhattan or London, it would be prohibitively time consuming and expensive. Each approach is valid and has its place. I think that's why our 3d team opted to employ both.

Below we illustrate the outcome of such an approach with Brighton and Swindon - the first digital cities within Virtual Earth in the UK:



See also a flythrough of Buffalo, one of the most recent additions to Virtual Earth (Music by Remergence):



The short term outlook is a further growth of digital cities from the two big players - Google and Microsoft. Smaller companies that have invested not inconsiderable sums in building models must be slightly worried but that is the nature of technology.

The construction of digital cities has reached a tipping point and within a couple of years we will be taking for granted the ability to fly into photorealisitc renditions of our urban areas - almost as much as we take for granted the ability to fly to anywhere on the earth, a mere 18 months since Google Earth was released.