It seams we are meant to work out the probability of after measuring the system with B and it becoming in state a0, getting the result 0 with the operator A. It's like a composed probability, something like that, one probability times the other.
The state at the beginning isn't told, I suppose...