NASA Logo, National Aeronautics and Space Administration

+NASA Home

+Ames Home

NASA OSTEM Fellowship Awards Autonomous Decision-Making Algorithms Research Project
Intelligent Systems Division Banner

NASA OSTEM Fellowship Awards Autonomous Decision-Making Algorithms Research Project

The NASA Office of Science, Technology, Engineering, and Mathematics (OSTEM) Fellowship has awarded the proposal titled, “Analysis and Evaluation of Autonomous Decision-Making Algorithms on Resource-Constrained Processors”. This proposal will fund the Ph.D. research of Timothy Darrah, working under Dr. Gautam Biswas, Cornelius Vanderbilt Professor of Engineering and Professor of Computer Science at Vanderbilt University. This fellowship award was issued in response to a proposed research idea submitted by the Autonomous Systems and Operations (ASO) project and was funded by ASO, managed by Dr. Jeremy Frank. Chetan Kulkarni and Ann Patterson Hine (Discovery and Systems Health technical area) wrote Letters of Support for this proposal. This research is also aligned with numerous projects being performed in the Intelligent Systems Division, including the Distributed Spacecraft Autonomy (DSA) project and battery prognostics work performed by the Discovery and Systems Health (DaSH) technical area under Aeronautics Research Mission Directorate (ARMD) projects.

BACKGROUND: This proposed research aims to fill the gap in space mission design: pairing a given decision-making algorithm for a given task to the processor that will be used to carry out that task on a given mission. Currently, space mission designers allocate thousands of hours of their team’s time to this evaluation process, which is error prone and offers no guarantee on solution quality. There are several factors that must be accounted for in the evaluation process as it relates to meeting minimum solution quality. These include hardware limitations such as processor speed and memory capacity, as well as algorithm limitations such as computational time and software complexity. While engineers and designers know the specifications of the various families of Central Processing Units (CPUs), and theoretical computer scientists know the computational complexity of the algorithms under study, ensuring that a selected algorithm will perform as expected on the selected processor under uncertain conditions is complex. Currently, sophisticated decision-making algorithms are not employed, either because they “don’t fit in the box” or the outcome can’t be guaranteed. The resulting analysis and framework will fill that gap, providing a guarantee on performance and solution quality, allowing better algorithms to be deployed, further enhancing mission capabilities, while saving thousands of design-time hours on future missions.

NASA PROGRAM FUNDING: Autonomous Systems and Operations (ASO) project, Human Exploration and Operations Mission Directorate (HEOMD)

POINT OF CONTACT: Jeremy Frank, jeremy.d.frank@nasa.gov

First Gov logo
NASA Logo - nasa.gov