The Pentagon146;s latest recruits: professors

The professors will receive grants of up to 600,000 per year for up to five years to engage in basic research8212; essentially a bet by the Pentagon that they will make a discovery that proves vital to maintaining the superiority of the US military.

8220;We do think great discoveries are likely,8221; said William S Rees Jr, the deputy undersecretary of defence for laboratories and basic sciences.

There was no shortage of applicant for the fellowships, perhaps because each fellow would be eligible for up to 3 million of grants. The field was cut to 20, and those applicants were asked to provide more details about their research plans and to meet with a panel of experts for interviews.

8220;I really didn8217;t know what to expect,8221; Rees said, adding that, as the proposals were evaluated, 8220;my knees were really staggered by the quality.8221;

In the end, six fellows were selected. Their research will be unclassified, but they will undergo background checks for a security clearance because the Pentagon wants to take them on tours of Defence Department laboratories, invite them to agency meetings and ask them to share their insights and knowledge with the department8217;s military and civilian leaders.

The Defence Department hopes that, over time, as more classes of fellows are selected, the grants will help forge relationships with some of the best minds in the country and enable the Pentagon to recruit them for advisory groups and study commissions.

Carey E Priebe, a professor in the applied mathematics and statistics department at Johns Hopkins University, is one of the fellows in the inaugural class. In nominating him for the fellowship, William R Brody, the president of Johns Hopkins, said Priebe8217;s research 8220;has tremendous potential to revolutionise the future of data analysis, for national security and other important applications.8221;

Priebe hopes to discover a way to make predictions from statistics calculated from multiple, disparate types of information, rather than the current practice of drawing inferences from a homogeneous source of information.

For example, Priebe8217;s idea would make it possible for a mathematical model and a computer programme to take information from scientific journals and link it to biographical information of individual researchers. For each researcher, the model would define his or her social network by tracking the scientific conferences they attend and whether their associates take jobs in labs with no apparent relation to their previous jobs.

The model, in theory, would piece together information not available in scientific journals and allow the Defence Department to predict innovations and avoid being technologically surprised by another nation or adversary.