Remember the SETI@Home project that was launched in 1999? It has been one of those pioneering projects where grid computing was brought to average consumers who would donate their computer hours for complex analysis of radio-signals from the space. The down-loadable software would receive data from the SETI server in the background and do calculations when the machine is idle at night or when you are away from it. Such grid computing way back then opened the doors for a new class of utility that would one day serve for complex stochastic simulations that could take years with limited computing power. A new possibility emerged!
Paying heed to this tremendous potential of volunteered grid computing across populations with desktops/laptos, MalariaControl.net hosted by AFRICA@HOME has launched a similar grid computing effort to model transmission dynamics and health effects on demographics with potential of getting infected by malaria. These simulations are intensive as they demand a huge set of grid points (human populations) with innumerable attributes that control the derivation of disease progression trajectory and other forecasting knowledge. The idea is simple. The Swiss Tropical Institute has developed a model for malaria epidemiology that uses the volunteer computing ability to calculate a credible analysis and a platform for evidence-based-treatment for malaria in Africa.
Again, the possibility of grid computing with volunteer computer hours is enormous. Further, to sweeten the deal, the architecture to support such efforts is FREE! The Berkeley Open Architecture for Network Computing (BOINC) is a openly available middleware that can be used to launch any such project. Another greatness of the BOINC framework is the support for both Microsoft Windows and Unix/Linux systems. Even though disease epidemiology studies for non-profit endeavors are great objectives for harnessing BOINC-like utility, the for-profit sector can also benefit from such volunteer contributions. Market research based on demographic behaviors and activity patterns across international borders is a demanding aspect of international business today. Stochastic marketing research models can leverage such grid computing efforts to run comprehensive analysis of product usage patterns. Furthermore, the volunteer attribute can be enriched by a monetary payment factor for hours of modeling usage of one’s computer too.
Picture: Courtesy AMREF