Technology's advance is central to economic growth and development. Furthermore, solutions for many of the planet's most pressing challenges --economic recovery, poverty reduction, climate change, sustainability-- require significant additions to society's technological toolkit. Yet, our ability to quantitatively model and forecast technological change is insufficient. The most common perspective on technological change in economics and management science--that it is a search on a space of technological possibilities--is more of a metaphor than a modeling framework. And the numerous case studies of particular innovations do not amount to a formal, quantitative and predictive theory. The project develops a formal methodology for describing the space of technological possibilities using a systematic, comparative analysis of U.S. Patent data spanning 220 years that explicitly accounts for the interdependencies of technologies at the most basic level. A major outcome of the project is a detailed "map" of technological capabilities in which potential innovation pathways are illuminated both visually and mathematically. The project mathematically formalizes and statistically analyzes this map in a form of network representation by utilizing and developing tools from mathematics, physics, biology and computer science. This quantitative and systematic approach is key to improving our ability to develop a predictive model that can inform decisions on public policy.
Among the fundamentalquestions the project answers are: (1) What is the appropriate "quantum" (fundamental unit of analysis) of technological capability required to understand invention activities? (2) How can we best construct a technology "map" that illustrates the complex interdependencies of these quanta, and that bridges from micro-scale dynamics to macro-scale patterns? (3) What characteristics define "keystone" (general purpose) technological capabilities in terms of their position within the ecosystem?
Broader impacts: The technology maps produced by the project will be available to the public and to policy makers. This information advances our ability to build a predictive model for technology forecasting and improves the ability of the US Patent and Trademark Office to index and monitor patterns in patent activities through time.
