Autonomous Urban Agents

Steve and I gave a well-attended presentation at MIT today as part of their Responsive Cities seminar series. We presented the "Santa Fe approach" to participatory modeling and decision making in urban planning applications, using an interactive 3-D sand table platform to demonstrate specific simulations employing autonomous agent technologies derived from research at the Santa Fe Institute.

Steve demonstrated the use of the Sand Table for:

• Wildfire simulations for emergency response training and incident command;
  
• Modeling of Venice’s boat circulation for dynamic traffic regulation involving stakeholders;
  
• Modeling pedestrian behavior to assist in the management of outdoor public space in Venice

The live demonstrations were followed by a discussion of the current direction of the research collaboration between Steve and me, illustrating possible future urban planning applications, based on interactive ambient computing platforms, such as:

• Interactive analysis and visualization of the (aggregate) impacts of urban plans
  
• Interactive user-assistance tools for permit applications in zoning and subdivision regulations
  

The presentation concluded with a discussion of an agent-based architecture for urban data management based on the City Knowledge concepts developed by Fabio at MIT-DUSP.

After the talk, Steve and I had a "deep" conversation with Raj Singh, fellow DUSPer who is now at the Open Geospatial Consortium (OGC) and Joe Ferreira, who advised Raj's dissertation at MIT-DUSP-UIS as well as mine. I hope we really tighten our collaboration to expand our web 3.0 concepts which lie at the cutting edge of urban information systems and planning...