Date: Nov 10, 2011 Source: SLAC Today (
click here to go to the source)
ML (Virtual Machine Language) has been used as the sequencing flight software on over a dozen JPL deep-space missions, most recently flying on Grail and Juno. As part of the NASA SBIR entitled Reactive Rendezvous and Docking Sequencer, VML version 3.0 is being enhanced to include object-oriented element organization, built-in queuing operations, and sophisticated matrix and vector operations. These improvements allow VML scripts to easily perform much of the work that formerly would have required a great deal of expensive flight software development to realize.
Autonomous slewing and tracking for the SBIR will make considerable use of the new VML features. Profiles generated by flight software will be managed using object-oriented VML data constructs executed in discrete time by the VML flight software. VML vector and matrix operations provide the ability to calculate and supply quaternions to the attitude controller in flight software which produces torque requests.
Using VML-based attitude planning components allows a reduction in flight software development effort, along with corresponding cost reductions. In addition, the direct management of the quaternions allows slewing and tracking to be tied in with sophisticated high-level VML state machines. These state machines provide autonomous management of spacecraft operations during critical tasks like a Mars Sample Return rendezvous and docking event. State machines created for autonomous science observations can also use this sort of attitude planning system, allowing heightened autonomy levels to reduce operations costs.
VML state machines cannot be considered merely sequences -- they are reactive logic constructs capable of autonomous decision making within a well-defined domain. The state machine approach enabled by VML 3.0 is progressing toward flight capability with a wide array of applicable mission activities.
