Automating Cassini instrument commanding for Saturn exploration

The NASA Cassini spacecraft is now in orbit about Saturn and in the prime phase of its mission to Saturn and the large moon Titan. Cassini was built and is managed by the NASA/California Institute of Technology Jet Propulsion Laboratory (JPL) in Pasadena, California. Unlike previous outer planet missions, the Cassini program employs a "distributed operations" approach to implementing the science-observing program. The associated ground system implementation is driven by this operations concept. Under this approach, the science instrument teams participate directly in planning and integrating the uplink command loads to the spacecraft, not only for commanding their own instrument states, but also for controlling the attitude of the spacecraft. This approach places responsibility for pointing control, as well as telemetry volume and power states, in the hands of the end users. This reduces the need for science-driven requirements to be passed from the science teams to uplink engineers, where resource constraints might result in compromises different from those the science teams might have chosen. Additionally, uplink engineers could potentially introduce errors in pointing, power management, or telemetry state with insufficient visibility by the science teams to recognize them. The distributed operations approach also introduces considerable complexity. The science instrument teams must gain a comprehensive understanding of the system level interactions and constraints and a mastery of the spacecraft command generation, integration, and validation tools normally controlled only by the uplink organization (JPL). Additionally, the science instrument teams must maintain a configuration management system usually required only within the uplink organization. The challenges represented by this expansion in the role of the science instrument teams has driven the development of many tools and procedures by those very lean teams, in order to minimize errors and schedule lapses that could cause such a distributed system to fail. This environment motivates the development of as much automation of the command generation and integration as possible, to keep the reliance on human work and re-work of command sequences to as low a level as possible