While much of Project Alpha’s research is driven by specific grants and problems, the team believes it is important to have a space in which to experiment with tools and techniques whose utility in non-proliferation analysis is not yet fully understood.
It is with this in mind that Project Alpha maintains a number of ‘innovation labs’ which are spaces in which to experiment with such tools and techniques.
Presently, Project Alpha has two such innovation labs.
Project Alpha undertakes machine learning to train algorithms to classify or predict. Our experiments in this area include
- Attempting to determine whether a scientific article is relevant to the nuclear fuel cycle
- Attempting to identify any proliferation-relevant items in an image
We are also continuing to explore other problems for which machine learning might be a solution.
We are experimenting with the use of VR in non-proliferation. We are specifically interested in the question of whether VR can help train safeguards inspectors. We are also interested in the question of whether VR can help us to identify proliferation-relevant sites from space.