Architecture

The semantic web holds tremendous promise for architects and designers. Tomorrow’s buildings will use sensory and schedule data to maximize efficiencies. Imagine a building that uses outside air to help regulate the temperature inside, and getting its wind and weather data from sensors miles away, so it can plan to be in the best position to harvest the incoming wind, sun, rain, air pressure, etc. Imagine buildings that will be able to change color and generate their own electricity – again they’ll do so with help from data coming in from several sources. Imagine buildings that can cooperate with each other, for example in handling large crowds or changes in traffic flow or flooded streets?

As with other industries, we must ask ourselves what will happen to the flow of information in the process of specifying, designing, building, and operating/maintaining our buildings. Will it stay about the same as it is today, only perhaps get faster and a bit better? Or will we start to use information in new ways that will have a profound impact on architecture, and the design of buildings fundamentally? If you read Pull, you’ll see that this shift is taking place in all industries, and it is already starting to affect architecture. Here are a few examples.

Home and office automation will be transformed by Internet-protocol-based information systems. Up until today, home/office automation was taken on by companies that had to provide everything, a total solution, not compatible with anyone else’s system. Over the years, these offerings have come and gone without much adoption, because they are too inflexible and based on a single vendor to provide everything. Now, companies are starting to spring up that use internet protocols and common standards to do electricity metering, energy management, track assets, and much more.

Virtual Design will start with requirements and a precise rendering of the location online, and those requirements will turn into sketches and potential solutions, which will then be tested and refined into prototypes, which then become a virtual model that can be tested and refined for engineering, heating and cooling, vibration resistance, lighting, traffic-flow, use-case analysis, maintenance, aging, costing, etc. Only when the model meets all the specifications does it then become a construction spec. By doing everything in one place, no one has to chase anything down. Instead of moving data files, we’ll keep the data in one place and let the tools and experts come and go.

Virtual Construction will integrate into the design space. Once you have a design finalized, you still need to have a construction plan that works within a given timeline and budget. All these elements will be brought into the model to specify what happens in the real world, so that as the building is constructed virtually in the same space where it was originally designed, the actual building is built. We’ll go back and forth between real life and the virtual space, updating and coordinating to account for changes as they are needed.

Bidding takes place in the same place, so the model is essentially the RFP. Different construction firms can come bid on various parts of the job and you’ll see all the bids in real-time, constantly adjusting automatically as the plans or price of materials changes.

Accounting also is centralized for the project, with money flowing in and out the same way information does.

Once the job is complete, operations uses the same info-space to manage everything in one place.

Maintenance is done using the same data. As things wear out or need refinishing, refurbishing, or replacing, each step is shown in the online space first, budgeted, bid, and executed virtually before anything happens in the real world.

You’ll find more thoughts and visions in chapter 9 of Pull.

Transitional Strategies

An example of home/office control is Arrayent Technologies, which produces an entire server-based package with connections to energy monitors and other devices.

An example of control automation is Proliphix, which offers proprietary HVAC control using web servers and a browser-based interface. It’s a start, but it’s a far cry from the open standards and plug-and-play systems we’ll have eventually.

In design, one interesting project is Google’s Sketchup environment, which lets people collaborate and share architectural models. It’s not the semantic solution of the future, but it’s a good step in the right direction.

On the proprietary side, Dassault Systemes is a leader in virtual design, and their web site is a rich resource for learning. They have excellent videos and tutorials. If you can imagine using Dassault tools in an open, shared, nonproprietary environment where everyone can not only bring in new tools and new talent, but also new formats and new datasets to keep things growing and improving, then you’re getting close to the fully semantic environment we’ll eventually have.

An exciting place to explore is Sensing Architecture, a site dedicated to new breakthroughs in architecture using new materials and new kinds of information systems.

Fully Semantic

If you know of anything in design or architecture that is fully semantic, please let us know.