Revolutionizing Space Exploration with Onboard Artificial Intelligence: A Case Study with NASA’s Perseverance DROID

NASA’s smart Martian DROID, Perseverance, is paving the way for a new era of space exploration with the use of Artificial Intelligence (AI). The onboard AI has the capability to make critical decisions autonomously, reducing the reliance on ground control and predefined procedures.

With an AI/ML Processing Unit onboard, the spacecraft can run an inference engine with pre-installed parameters that can be updated through telecommands. This allows for more efficient decision-making and data processing, without the need for constant communication with Earth.

One of the key advantages of having AI onboard is the ability to reduce the size of neural networks, making data transmission more cost-effective. By using reduced-precision and bare-minimum neural networks, the spacecraft can upload valuable scientific data more quickly and efficiently.

In a recent study focusing on NASA’s Magnetospheric MultiScale (MMS) mission, researchers demonstrated how AI can be used to classify data in Earth’s magnetosphere. By selectively downlinking high-value data and triggering burst-mode data collection in regions of interest, the AI onboard Perseverance is able to optimize data transmission and collection.

The study also showed how neural networks can be reduced in size by up to 98.9% without sacrificing accuracy, by using lower-precision formats to represent network parameters. This not only reduces the overall size of the model but also decreases the cost of data transmission.

Overall, the integration of AI into space exploration missions like Perseverance is revolutionizing the way we collect and analyze data in space. With onboard AI capabilities, spacecraft can operate more efficiently and effectively, leading to groundbreaking discoveries and advancements in our understanding of the universe.