Psyche: Innovations in Development of Planning and Sequencing Systems

NASA Jet Propulsion Laboratory’s (JPL) mission Psyche will be the first of its kind in many ways: the first to explore a metal asteroid, the first to fly a Hall Effect Thruster (HET) in deep space, and the first to demonstrate Deep Space Optical Communication (DSOC). Challenges presented by firsts of these nature require powerful, yet agile ground and onboard systems to be addressed in a timely and effective manner. This paper examines the ways that innovations in Psyche’s planning and sequencing systems address both common challenges faced by many missions and unique challenges arising from first ever-in-flight activities. Psyche will be the first mission to operate using JPL’s new common core flight software Flight Core Product Line (FCPL) and make use of capabilities such as real-time telemetry reactivity, multi-level logic, and complex math operations which enable new levels of onboard autonomy. Augmentation of FCPL’s new capabilities through a user-friendly abstraction layer in ground software, an effort undertaken in conjunction with NASA’s Europa Clipper mission, enables even more complexity in reusable sequences, as well as streamlines and simplifies the ground operation and review process. The combination of increased onboard capability in the reusable sequences with Psyche’s agile planning and scheduling ground software allows Psyche to quickly develop new solutions to complex problems and represents an advancement from the approach of previous missions. An example of one of these solutions that takes full advantage of ground and onboard capabilities is presented: a unique momentum management strategy that addresses the challenge of the "swirl torque" that arises from the use of HETs in space. In order to address the challenges of testing these advanced reusable sequences, Psyche developed a robust and comprehensive automated regression testing framework and an intelligent sequence branch analysis tool that enables quick and reliable reusable sequence re-testing against new Flight Rules and constraints that are identified, resulting in significant risk reduction and cost-saving over manual approaches taken by previous missions. Analysis of these tools will prove useful to future missions that develop blocks, reusable sequences, or other onboard behaviors. A multi-mission process to parse command dictionaries into a set of Python command classes allows Psyche’s sequencers to develop sequences directly in Python, a significant improvement over the previous approach of sequencing in Domain Specific Languages (DSLs) This approach of developing Python command classes provides the ability to develop sequences in any Integrated Development Environment of their choice, and results in significant cost savings by avoiding the need to develop sequence editing graphical interfaces, as have been developed for missions in the past. Benefits, drawbacks, and work to go for each of these innovations are discussed and lessons from Psyche’s development phase are provided.