I think the main issue is the ability to follow the information through the information systems. They identify high-priority repairs that are required, yet the way they organize that information is not the same way that they organize the information when they actually conduct the repair work. It's very difficult to trace it through.
Moreover, there doesn't seem to be a very good vetting process of what actually is a high priority. When we looked at the number of repairs, they came back and said that many of them had been dropped because they weren't a high priority. It's a question of the management of that list and then of follow-through. If the list exists, they should ask whether the work has actually been done, and someone should be monitoring that. They recognize the problem and have agreed to fix it.