JSON-Based Sequencing For Deep Space Missions

Robotic deep space exploration missions, such asthose commonly operated by the NASA Jet PropulsionLaboratory, rely on “sequences” as the primary means ofnominal commanding. Each sequence is a unique file containingone or more commands that are timed to execute at some pointin the future – in this way, the operations team on the groundcan maintain control of the spacecraft through periods when itis out of contact, or when speed-of-light induced delay makesreal-time commanding impractical. Sequences are sent to aspacecraft in a binary format that is interpretable to theonboard computer, but must be developed on the ground in asource format which can be understood by human operators.The desired characteristics of a sequence source format areoften competing; it must be easily written and understood byhumans, but it also should be serializable to support automatedgeneration and review of sequences. The syntax and semanticsof the format must be unambiguously specified in a way that canbe communicated to disparate instrument and subsystemoperations teams that need to write and deliver sequences. Anecosystem of supporting software tools must be also be providedto operations teams to enable authoring, editing, simulation, andreview of sequences in the chosen format before they are sent tothe spacecraft. An inappropriate or poorly-designed sequencesource format has the potential to increase operations cost,hinder ground system automation efforts, confuse operationspersonnel, and lead to command errors during flight. TheEuropa Clipper and Psyche projects jointly developed theSeqJSON language as a new source format for spacecraftsequences, replacing legacy formats such as Spacecraft ActivitySequence File (SASF) and Spacecraft Activity Type File (SATF)that have been in use on deep space missions operated by the JetPropulsion Laboratory for decades. It is based on the JavaScriptObject Notation (JSON) format – a lightweight datainterchangeformat that is widely compatible with existingsoftware tools and programming languages. SeqJSONsequences are easy for humans to read, but also easy forsoftware to ingest, modify, and output. The syntax of SeqJSONsequences is fully specified by a JSON schema, which can beprovided to remote teams and used at their home institutions foron-premises sequence validation. The SeqJSON language wasdeveloped with Europa Clipper and Psyche onboard flightsoftware in mind, but it is applicable to any spacecraft that usestime-ordered sequences for commanding. This paper discussesthe SeqJSON language, how it differs from and improves uponlegacy formats, the operations paradigms it supports, thesoftware tools that facilitate its use, and the way it is being usedtoday.