NASA Funds Robotic Innovation: Pushing Boundaries of Exploration
New projects aim to make robots smarter, more reliable for challenging missions in space
The National Aeronautics and Space Administration (NASA) has awarded grants to several research teams, propelling advancements in robotic technology crucial for future space exploration. These projects, tackling a variety of areas from icy moons to the depths of the ocean, aim to create more intelligent and resilient robots capable of handling complex tasks autonomously in extreme environments.
One key focus is developing robots that can "think" more flexibly and adapt to unexpected challenges. Jonathan Bohren of Honeybee Robotics, for example, is leading a team working on “Stochastic PLEXIL," a system that uses machine learning techniques to give robots the ability to make decisions in the face of uncertainty. "Extended PLEXIL with stochastic decision-making capabilities by employing reinforcement learning techniques," Bohren explained. This could be crucial for missions to icy moons like Europa where unpredictable terrain and conditions pose significant challenges.
Pooyan Jamshidi from the University of South Carolina is taking a different approach, focusing on making robots more reliable. His team is developing tools for “Resource Adaptive Software Purpose-Built for Extraordinary Robotic Research Yields (RASPBERRY SI)," a project meant to pinpoint the root causes of malfunctions and improve the robot’s ability to self-diagnose and recover.
[Image of a robot arm on an icy surface. Credit: NASA]
The COLDTech projects specifically address the challenges of exploring extremely cold environments, like those found on the moon. Eric Dixon of Lockheed Martin is leading a team integrating a model of JPL’s Cold Operable Lunar Deployable Arm (COLDarm) into a simulated environment, enabling the robot to learn and adapt to ice buildup and other potential issues.
Meanwhile, Jay McMahon from the University of Colorado is developing a system called REASON-RECOURSE, which uses automated planning and formal methods to maximize a lander’s scientific output while minimizing communication with Earth.
These advancements promise more exciting possibilities for robotic exploration. Imagine robots capable of autonomously navigating icy moons, analyzing geological formations, and even collecting samples for analysis, all without constant human intervention.
As Joel Burdick from Caltech aptly put it: “Developed autonomous 1) detection and identification of off-nominal conditions and procedures for recovery from those conditions, and 2) sample site selection.”
These projects represent just a glimpse into the future of robotic exploration, a future where these intelligent machines will play a critical role in unlocking the secrets of our solar system and beyond.
