Taxonomy for Applied Human-Machine Teaming for Space System Domains

Future long-duration space missions will require significant use of intelligent and autonomous systems to support crew self-sufficiency. These future missions to the Moon and Mars will require significant infrastructure reliant on the capabilities of autonomous systems to support orbit, surface, and science operations. Furthermore, these missions will involve prolonged periods of crew isolation, confinement, and exposure to the extreme environments of space which presents novel challenges for teams operating in these conditions. Current systems require significant involvement and oversight by human operators, pulling them away from mission. Technology with greater autonomy capabilities does not necessarily lead to better mission outcomes. Instead, the ability to have humans and intelligent machine systems (autonomous, AI, and robotic systems) working collaboratively, and even as part of a team will be a critical enabler for supporting complex mission operations within the space domain. For the promises of Human-Machine Teaming (HMT) to be realized, a shift of focus is needed from greater operational autonomy (operating with less human involvement) toward greater autonomous teaming capabilities (operating with higher productivity with a human teammate). It is important to note that higher levels of autonomy alone, while necessary, will not drive us towards better HMTs. Additionally, an autonomy-centric design perspective won’t get us to the advanced HMT capabilities envisioned for the future. Inevitably, crews will have to operate under conditions that are stressful for effective team performance such as high workload, reduced sleep, extended isolation from family, and communication delays. Higher levels of dependency by crew on their autonomous teammates under these conditions necessitates that these systems be trustable and reliable. A novel taxonomy for multiple levels of HMT is proposed describing incremental levels of HMT to help identify mission needs-based requirements for different types of human-machine teams. This taxonomy provides structure for design decisions and drivers that can be tailored based on the nature and extent of desired teaming capabilities.