NASA’s LASSIE robot dog trains to take on the Moon

The LASSIE (Legged Autonomous Surface Science In Analogue Environments) project investigates the capabilities of legged roving platforms in assessing the geotechnical properties of crusted and icy surface regolith. Utilising the leg motors, these platforms measure these properties and use the data to autonomously revise their science operations plans.

This research is currently being conducted in environments that simulate Martian and lunar conditions in Mount Hood, Oregon, as well as within laboratory settings. The icy and gravelly terrain mimics what scientists expect to see on the lunar and eventually Martian surfaces.

The project is led by Feifei Qian, a Georgia Tech Alumna and Assistant Professor at the USC Viterbi School of Engineering. She is working alongside Frances Rivera-Hernández, another Assistant Professor from the School of Earth and Atmospheric Sciences at Georgia Tech. This initiative represents a joint effort among prestigious institutions including NASA, USC, Texas A&M University, Oregon State University, Temple University, and the University of Pennsylvania.

Spirit, a four-legged robot, is conducting tests on the challenging terrains of Mount Hood's Palmer Glacier, replicating the severe conditions of the moon and other celestial bodies. This robot is engineered to modify its locomotion techniques in real time, an essential skill for traversing the unpredictable landscapes of other planets. As it moves across dirt, rocky slopes, and slushy snow, Spirit gathers data that will guide the design of future robotic explorers, ensuring their capability to navigate extraterrestrial terrains.

The robot's ability to detect the mechanical resistance of various surfaces and instantly adapt its walking strategy is pivotal to its training. This technology replicates the tactile feedback humans perceive while walking on uneven terrain, offering vital insights into the surface compositions and movement dynamics on other planets.

Why robot dogs over traditional rovers?

Planetary surface exploration missions typically follow a pre-set plan, which can restrict their ability to respond effectively to unforeseen variations in terrain properties. This limitation affects their capacity to modify locomotion or sampling approaches and to autonomously pinpoint and adapt to scientifically important findings. A particular challenge is the difficulty in measuring and responding to unexpected shifts in regolith properties, which can adversely affect mission operations. Both the Mars Exploration Rover Spirit and the InSight lander encountered issues related to the geotechnical properties of regolith. However, developments in legged robotic platforms offer promise for overcoming these obstacles by enhancing sensitivity to changes in surface properties.