2017 AEC Technology Hackathon

Kyle Martin   •   over 8 years ago

Reactionary Design Optimization for Urban Master Planning

CONCEPT DESCRIPTION:
The traditional master planning process is time consuming and only produces a singular outcome containing a mixture of parcel sizes, FAR, use types, program, design standards, etc. Once buildings are locked in for specific parcels, the remaining portions of the master plan must be evaluated to assess reduction in available assets and mixture. We have an idea for a computational approach to urban planning studies that explores a multitude of possible schemes and strives for balance across entire site.

PROCESS:
- capitalize on an existing master plan with numerous available assets
- set up tools, test on a handful of initial parcels > test and reconfigure
- process remaining parcels and further tune mechanism
- generate design space for each parcel based on provided constraints
- gather all options into comprehensive database
- run machine learning process on database of options > test and tune as needed
- export premium design schemes
- re-run optimization as parcel designs are locked in and evaluate reaction across site
- examine analytics such as: construction cost, energy efficiency, tenant/lease breakdown, etc.
- create a web interface and user experience for accessing and manipulating outcomes

KEY FOCUSES:
1. Rule Sets - how do we nicely code up the constraints?
2. Analytics - what are the takeaways from resulting dataset?

GOALS:
- learn about various pertinent feasibility factors
- gain a deeper understanding of the tools and methods available for rapid iteration and optimization
- develop a working concept using coding or other means
- have each team member walk away from the experience with at least one new skill