CHAPTER 10 DAY IN THE LIFE: PREPARING FOR THE UNEXPECTED 180 (e.g., the spacewalk) continuing as scheduled. Flight-specific simulations are focused on a particular upcoming event wherein the actual team members who will execute the event train together. Generic simulations are populated with people at various levels of training. Personal photo courtesy of Robert Dempsey Figure 3. The training team, led by CTO William Frank (green shirt) in their own Mission Control-like facility. A computer generates the fake ISS data that the flight controllers see in Mission Control. The training team is also looking at these fake data. However, unlike the flight controllers, the instructors can inject failures into the simulation at the stroke a mouse click. For example, they can make a pump suddenly overheat and fail, or make a computer start generating erroneous messages. Failure scenarios can become fairly complicated. One or two system failures may not be much of an impact to the operations however, they can combine to cause significant constraints or vulnerabilities to an additional failure. For example, consider a module that has two internal thermal control loops and two power strings: A and B. Thermal pump A is on power string A thermal pump B is on power string B. If power string A fails, string B is used. Now, thermal pump B becomes more critical. Should anything happen with the thermal loop (e.g., the pump fails) or if power string B is interrupted, all cooling to that module will be lost. In this type of scenario, the team will try to anticipate the next-worse failure after the first failure. Thus, when string A goes down, the team will try to anticipate what to do to best protect the vehicle for the next failure. In this example, the team might get a power jumper ready to reactivate pump A by plugging it into the B system. The training team spends a significant amount of time learning the systems, the timeline, and the objectives in preparation for a simulation. This research may involve going into the simulator, testing failures, and seeing how the software responds. Not only does this help lead to a realistic simulation, on many occasion real software bugs have been caught before being loaded onto the actual vehicle. Several weeks in advance of a simulation, the CTO will lead development of a script for the run. The script defines what failures will occur, and when and what the training team anticipates the flight control team will do in response. Sometimes, however, the flight control team will make a decision, either in error or on purpose, and choose a different response than anticipated. In this case, the training team will conduct