Long-Duration, Risk-Aware, Goal-Directed Adaptive Sampling for Autonomous Vehicle Exploration

As lunar operations increase in support of Artemis missions, there will be more uncrewed vehicles operating on the surface of the moon. The Artemis Plan intends to enhance science investigations on the lunar surface by expanding the use of mobility systems. In order for NASA to execute a significant percentage of its decadal science survey and Artemis Plan priorities, it is essential that NASA limit growth of the operational costs for these mobile robotic missions. This desire to limit operational costs reaches well beyond lunar missions, for example, to lowering the operational cost of New Frontiers 5 Missions (Section 5.6.1 of the draft AO). To reduce the need for human intervention required in autonomous exploration, my research will develop innovative technology such that scientists can directly pose queries and direct observations to the rover through goal-directed adaptive sampling, as well as expand the the duration of autonomous operation through risk-aware long-duration planning. These capabilities are novel and do not currently exist beyond TRL 1 or 2. I propose to pursue research on autonomous methods for long-duration adaptive sampling that are simultaneously query-directed and goal-directed while being risk-aware. Specifically I use these terms to mean the following: Query-Directed: Systems that use an information query from scientists to plan the path of their observations. Goal-Directed: Systems that can simultaneously pursue information goals from queries and incorporates state-based goals. Risk-Aware: Systems that can assess and minimize the risks that they take with respect to their own safety and to state-based and information-based goals. The current state of the art is query-directed and risk-aware. My research will extend this work to account for long-duration operation, as well as incorporating state-based goals into adaptive sampling planning. This method will use a receding horizon planner that can plan over chance-constrained state goals, and Bayesian networks to model the information gain. This research aligns with NASA's STMD strategic framework EXPLORE thrust in the domain of Autonomous Systems and Robotics (ASR). One of the ASR technology objectives is the have efficient on-board autonomy for continuous surface operations with cost- effective mission control. To accomplish this, autonomous vehicles will need to operate with significantly less human oversight. There are two potential avenues to reducing this intervention: reducing the number of operators that mediate between scientists and vehicle operation and extending the periods during which the vehicle can operate autonomously.

This is early/mid-stage (TRL 2) — the most realistic path in is NASA SBIR/STTR, which funds small businesses and research institutions to develop technology aligned with NASA's needs (equity-free, phased funding). Check whether a current SBIR/STTR solicitation topic overlaps with this project's technology area, or contact the project directly (above) to ask.

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.