Automated Scheduling of Science Activities for Titan Encounters by Cassini

Abstract

In an effort to demonstrate the efficacy of automated planning and scheduling techniques for large missions, we have adapted ASPEN (Activity Scheduling and Planning Environment) [1] and CLASP (Compressed Large-scale Activity Scheduling and Planning)[2] to the domain of scheduling high-level science goals into conflict-free operations plans for Titan encounters by the Cassini spacecraft.The Cassini mission is a cooperative undertaking between NASA ,ESA and ASI and has been in orbit for nine years, returning a wealth of scientific data from Titan and the Saturnian system. Previous work has documented how the Cassini operations team has historically produced operation plans [3]. Automated techniques might be used to increase the responsiveness of science operations planning, reduce the costs of such planning, and also provide insight to mission design decisions during mission planning.With respect to reducing cost, science plans entail a significant period of a project's prime mission. The Titan Orbital Science Team (TOST) spent considerable efforts to integrate the science plans in the +/-20 hours around the targeted flybys [4]. The Titan team produced master timelines for each flyby, identifying prime science observations and allocating control of the spacecraft attitude to specific instrument teams. This effort required dozens of science team members, and substantial engineering team support over many years in a highly contentious, intensely concentrated effort. Our approach is to quickly generate these plans and then iterate with science teams in the loop to converge on an overall plan that meets science objectives and has buy in from all teams. This iteration is necessary as we need to balance the limited shared resources (particularly pointing) of the spacecraft between the disciplines (Titan Interior, Surface, Atmosphere, and Magnetospheric Interaction) and the 12 instruments of the Cassini Orbiter (which include 4 optical remote sensing, 1 radio science, 1 RADAR, and 6 fields and particles instruments). Once a final plan is established, certain events can still lead to necessary adjustments -e.g. changes to the nominal trajectory, a failed instrument, DSN losses/changes, spacecraft safing or ground events [5]. In such cases, ASPEN and CLASP can be used to assess the impact and replan relatively quickly.The adaptation of ASPEN and CLASP for Titan flyby planning focused on representing instrument contention and science team scoring. We document the various states and resources used to model these, as well as the activity descriptions that decompose high-level science goals into detailed, conflict-free operations plans.