With no GPS in space, how can a rover know its exact location on a lunar surface?

In this Chip Chap podcast, Phil Ludivig, rover navigation engineer with iSpace, Inc.* joins Shashi Jain, innovation manager at Intel, to talk about research that applied AI to one of the biggest challenges in space exploration.

Ludivig and Jain, along with other researchers, came together at NASA Frontier Development Lab (NASA FDL) to tackle questions facing NASA and the commercial space industry. Their team took on one of the most fundamental – and answered it a highly inventive way.

Starting with a game engine, the team created a simulated lunar environment to train an AI algorithm that produced the ground truth needed for machine learning. Next, they created synthetic images, called reprojections, from cameras mounted on a rover. AI matched reprojected images to actual orbital images, figuring out terrain features that made sense.

The team used Intel® AI DevCloud for inference, an Intel Core™ i7+ PC and Intel Xeon® Scalable processor-based server for synthetic training data generation, and Google* Cloud Platform for training.

The same technique can be applied anywhere, on Mars or even areas of Earth where GPS is out of reach.

For details about this and other NASA FDL projects, visit frontierdevlopmentlab.org.

Information about Phil Ludivig’s organization is online at ispace-inc.com.

More about AI at Intel is available at intel.com/ai.

Intel technologies' features and benefits depend on system configuration and may require enabled hardware, software or service activation. Performance varies depending on system configuration No product or component can be absolutely secure. Check with your system manufacturer or retailer or learn more at intel.com.
Intel, the Intel logo, Core, and Xeon are trademarks of Intel Corporation or its subsidiaries in the U.S. and/or other countries.

*Other names and brands may be claimed as the property of others.

© Intel Corporation