Project Summary

The proposed research aims to develop Pro-Opt, a computational optimizer that project managers can use to identify near-optimal designs for their future project organizations.  Pro-Opt uses Genetic Programming, an evolutionary computing method developed by Professor John Koza in Stanford’s CS department, to iteratively optimize predictions from the Virtual Design Team (VDT) organizational analysis software developed at CIFE.  Our first year research on this seed grant has demonstrated that integrating evolutionary computing methods with computational analysis tools for this domain holds exceptional promise.  This bodes well for the broad applicability of evolutionary computational optimizers to a number of other CIFE analysis tools, such as tools for prediction of hazards or costs in construction work processes using extensions of 4D-CAD developed by several of Professor Fischer’s students, or the “Perceptor” analysis tools developed by John Haymaker that will be used to address a variety of sustainability concerns in the future.