Vehicle Platform    Mission Specific Payload and Sensors    Mission Strategy

Mission Strategy

At the highest level, the operational AI is simply a state machine, with one state for each major mission task.  Once a task is completed, we proceed to the next state. 

The initial state is very simple, consisting of a dead reckoning through the validation gate.  Once we pass through the gate, we begin a search phase to locate the indicator arrows and search for the active light in the array.

Upon locating the indicator arrows and detecting the active light, we extract a heading from the arrow in question, and also use this heading to determine the frequency of interest for the hydrophone array. 

Initial navigation to the target is accomplished through the extracted heading.  During the navigation phase, additional feedback is also provided by the hydrophone array and the vision processing subsystem.  This allows for a robust mechanism that makes use of the acoustic pingers at longer ranges, but switches to a visual system as we approach the target assembly for greater precision.

During the final phase, we use the two cameras to navigate the submarine directly over the target array.  The downward firing altimeter is also employed for accurate depth determination, as the beam spread on the DVL makes it possible to miss the target when it is right beneath the submarine.  Once we are aligned with the target, a bank of electromagnets drops two markers.  The submarine then surfaces, terminating the mission.

The AI structure at the highest level is relatively straightforward, as the shared memory structure allows for the independent development and testing of the various software components necessary, allowing the high-level AI to merely start and coordinate the various components as necessary in any given phase.

<= Back: Mission Specific Sensors and Payload