There's a history of fusion projects that are characterized by cost overruns and delays. How do you minimize the chance of this? You have to build great teams, as one of them. The other way to make it faster is modularize the problem or parse the problem.