Robust, Explainable Autonomy for Scientific Icy Moon Operations

Our goal is to address the needs of future ocean worlds lander missions that will require reliable operation of a robotic arm and its science instruments. This effort leverages autonomy technologies developed specifically for implementation on-board, to credibly and uniquely address COLDTech goals:

Increase productivity of surface science operations with reduced dependence on operator input: We encode task flexibility and inform system planning to continue operating even in the presence of faults and unexpected conditions and events, alleviating schedule impact and reducing operator intervention. Our objective is to reduce the frequency of needed ground contacts by a factor of 10.

Key to this is our approach to: Enable autonomous adaptation to faults, including those due to unknown and uncertain environments: We use a unique anomaly detection technique based solely on models of expected behavior, enabling comprehensive anomaly identification, the information needed for re-planning, reconfiguration, or re-calibration, as appropriate. We will show autonomous mitigation of at least half of operations-impeding anomalies that have interrupted previous planetary missions. We build trust in this capability by having the technique explain its reasoning and conclusions throughout a test program conducted in the OceanWATERS and OWLAT testbeds

Our Caltech-led team will demonstrate function- and system-level capabilities, focusing on anomaly identification and adaptation through autonomous recalibration and synchronous model updates of both the hardware and, crucially, its environment. Our methodology is to (1) develop a prototype of the closed-loop, fail-operational monitor and control system; (2) produce behavior and diagnostic models of representative hardware and the environment suitable for on-board execution as part of a complete system; (3) integrate and demonstrate the prototype as adapted to the two testbeds. The approach extends previous autonomy efforts that have been proven flight certifiable, offering a viable path to infusion and a high return on investment by harvesting the following:

Adaptable algorithms for robotic mobility and manipulation, developed for robust exploration and manipulation of unknown objects (Caltech)

Reactive autonomous planning technologies and a system-level autonomy framework to robustly complete complex activities following anomalies and degradations (JPL)

Comprehensive on-board physics-based anomaly detection and diagnosis techniques to assess system and environment states even in novel conditions (Okean) Our end result will provide autonomous monitoring and control of robotic systems and the environment in which they operate, capable of adaptation to unexpected conditions. This enables a systematic approach to drive appropriate "fail-operational" responses through robust, explainable, self-reconfiguration and recalibration capabilities encapsulated in critical robotic systems.

This work is motivated by a recognition of the persistent problem of error-tolerant task execution such as reliable operation of a robot arm and science instruments. This is especially crucial for contact science involving complex environmental interactions. In operations, when manipulator and instrument operations experience anomalies, the default is to suspend science operations and await ground intervention. While this does preserve spacecraft health and the potential for future observations, it may severely limit science throughput, even becoming the main limiting factor if spacecraft life is limited or observation opportunities are few. Studies of MSL operation concluded that such soft failures may reduce useful science throughput by a factor of five. On an icy moon sampling mission, such inefficiency would lose precious science; our goal is to provide trustworthy, robust autonomy for landed missions.

This is a mature technology (TRL 7+) — the realistic path in is usually NASA's Technology Transfer Program: licensing an existing NASA patent, or a Space Act Agreement to use NASA facilities/expertise directly. NASA also runs a startup licensing program with no upfront fee for companies formed to commercialize a specific NASA technology.

None of these are guaranteed paths for this specific project — TechPort itself doesn't have an "apply" button. Reaching out to the contact(s) above with a specific question is usually the fastest way to find out what's actually open.