Description: Deep-space missions will require spacecraft to overcome unknown problems and unexpected system failures in novel regimes without the support of Earth-based communication or resources. Reliable autonomous systems are needed to assist crew with managing and executing missions, and responding to unanticipated system failures and anomalies, thereby reducing cognitive burden of the crew and time spent on troubleshooting and preventative maintenance. VISIMO proposes an autonomous agent Cognitive Architecture (CA) to identify and mitigate system failures on deep-space missions. The CA will locate problems, notify the most relevant spacecraft system or human team, and map solutions to unexpected problems. VISIMO has partnered with Hewlett Packard Enterprise (HPE), who has offered their Spaceborne Computer-2, located aboard the ISS, as a venue for Phase II testing and validation of VISIMO’s solution in relevant operational contexts, allowing VISIMO access to the International Space Station (ISS). VISIMO will refer to the system as GRaceful Architecture for Mitigation of System failures, or GRAMS, which highlights the important properties of the architecture—that it is designed using principles of Graceful Extensibility; that its emergent networks allow it to identify “unknown unknowns;” and that the system not only identifies possible failures, but also provides mitigations to the appropriate recipient. The CA connects to the ISS, Moon to Mars, and Advanced Exploration Systems (AES) missions. The proposed Phase I application will be tested against a realistic model of the Atmosphere Control and Supply (ACS) subsystem of an Environmental Control and Life Support System (ECLSS), such as that used on the ISS, to confidently determine feasibility and reliability.

Benefits: GRAMS will identify and mitigate spacecraft system failures on deep-space missions by locating problems, mapping solutions for dynamic environments, and detailing potential consequences of mitigation actions. The software will thereby reduce the cognitive burden on crew members and time spent on preventative maintenance, maximizing crew time for research and scientific exploration.

GRAMS supports improved autonomy in situations that require high levels of assurance in dynamic environments, including certification of land, air, and sea autonomous vehicles. The tool can also be used for risk and failure monitoring of non-vehicle systems like the management of nuclear and chemical plants through alerting personnel of upcoming risks and engaging automatic safety systems.