The term I picked up on was "universities". Universities do not get involved in contracts. They live in the world of DAPRA grants, typically at level 6.1 and rarely 6.2 (old terms for "basic research" and "transition technology"). DARPA's grants are often blue-sky stuff: "Imagine that we had sharks with laser beams on their heads. How could we use them?" DARPA is encouraged to think forward even when the technology support doesn't exist. That's what the first "A" in DARPA stands for. So just because this is a system for doing real-time video surveillance (which is fairly common topicwise), and the blue-sky example is a keyhole satellite, doesn't at all mean that there's a keyhole satellite which enables real-time video surveillance. It just means the project manager is being encouraged to dream big.
Heuristics are bug ridden by definition. If they didn't have bugs, then they'd be algorithms.