The Department of Energy is running a competition for ideas on preventing nuclear proliferation using heterogeneous data. To quote James Kornell who is running the competition:
Nonproliferation is, it seems to me, a particularly messy problem. It can be violated at both the state and the non-state level, and it involves technical, cultural, political, military, and sophisticated deception aspects. In addition, it’s urgent: a couple of the long-timers I know who are still active believe it’s a near-inevitability we’ll see a nuclear device detonated in a major city in the next 15-20 years. They earnestly hope they’re wrong; but as one of them put it, ‘you could build one in your garage.'
So — heterogeneous data with multiple levels of certainty and consistency, incomplete data, intentionally deceptive data, ambiguous signal, non-unique signatures, and the few physical signals trivially easy to hide. Not to mention international law and the difficulty of validation. We need to recruit netsci researchers to this problem space.
The formal call is linked here. Entries are due by July 15. The prize is $2,000.