Multi-Rover Integrated Science Understanding System (MISUS) is a computer program designed to coordinate the activities of multiple small, instrumented robotic vehicles (rovers) engaged in autonomous scientific exploration of the surface of Mars. MISUS includes a component that utilizes machine-learning clustering methods to analyze scientific data (principally, image and spectral features of rocks) and, on the basis of analyses, to select new scientific activities. MISUS also includes a distributed-planning-and-scheduling component that determines the rover activities needed to achieve scientific goals, partly on the basis of initial rover conditions and an input set of goals. Plans are updated on the basis of the results of the scientific analyses and current information on the execution of commands and utilization of resources. Planning is distributed among the individual rovers, each rover being responsible for planning its own activities. A central planning system is responsible for dividing up the goals among the individual rovers in a fashion that minimizes the total time of traversal of all rovers. The software as described thus far is also integrated with a simulation program that simulates multiple-rover scientific operations on Mars-like terrain.
This program was written by Tara Estlin, Alexander Gray, Darren Mutz, Ashley Davies, Eric Mjolsness, Gregg Rabideau, John Lou, Rebecca Castaño, Steve Chien, and Tobias Mann of Caltech for NASA’s Jet Propulsion Laboratory.
This software is available for commercial licensing. Please contact Don Hart of the California Institute of Technology at (818) 393- 3425. Refer to NPO-30201.
This Brief includes a Technical Support Package (TSP).
Software for Scientific Exploration by Multiple Rovers
(reference NPO-30201) is currently available for download from the TSP library.
Don't have an account?
Overview
The document outlines the Multi-Rover Integrated Science Understanding System (MISUS), a sophisticated software program developed for coordinating multiple small robotic rovers engaged in autonomous scientific exploration on Mars. Created by a team from Caltech for NASA's Jet Propulsion Laboratory, MISUS integrates machine learning, planning, and scheduling techniques to enhance the efficiency and effectiveness of planetary science missions.
MISUS employs machine-learning clustering methods to analyze scientific data, primarily focusing on image and spectral features of Martian rocks. This analysis helps classify different rock types and informs the selection of new scientific activities for the rovers. The system's planning and scheduling component determines the necessary activities for each rover to achieve specific scientific goals, taking into account the initial conditions of the rovers and the overall mission objectives.
A key feature of MISUS is its distributed planning capability, where each rover is responsible for planning its own activities based on the central system's directives. This approach minimizes the total traversal time for all rovers, optimizing their movements across the Martian terrain. The software is also integrated with a simulation program that models multiple-rover operations in Mars-like environments, allowing for testing and refinement of strategies before actual deployment.
The document emphasizes the importance of using larger sets of rovers for future missions to maximize scientific return and enable complex science activities. By facilitating autonomous multi-rover behavior, MISUS aims to improve the analysis of science data and enhance the decision-making process regarding new observations and actions.
Additionally, the document includes a reference to a publication detailing the integrated system, presented at the Sixteenth National Conference of Artificial Intelligence in 1999. It also clarifies that the software is available for commercial licensing, with contact information provided for inquiries.
Overall, the MISUS framework represents a significant advancement in robotic exploration technology, promising to enhance the capabilities of future Mars missions and potentially other planetary explorations. The integration of machine learning with autonomous planning and scheduling marks a pivotal step toward achieving more ambitious scientific goals in space exploration.
