Obsessed with technology?
Subscribe to the latest tech news as well as exciting promotions from us and our partners!
By subscribing, you indicate that you have read & understood the SPH's Privacy Policy and PDPA Statement.
News
News Categories

Intel powers first satellite with AI on board

By Raymond Lau - on 26 Oct 2020, 9:53am

Intel powers first satellite with AI on board

(Image: Intel)

On September 2, artificial intelligence made its way into space on an experimental cereal box-sized satellite named PhiSat-1.

“What could possibly go wrong?”, asks my friend Dave.  

Armed with a hyperspectral-thermal camera and onboard AI processing thanks to an Intel Movidius Myriad 2 VPU (visual processing unit), the PhiSat-1 is on a mission to monitor polar ice and soil moisture and to test inter-satellite communication systems. (The Intel Movidius Myriad 2 VPU was first seen in Intel's Compute Stick 2 offering.)

AI is enlisted to help with deciphering whether the high fidelity photos taken by the satellite are worth sending down to earth. For example, if the AI can filter out and discard useless images of clouds, the satellite can conserve about 30 percent of bandwidth and ensure only useful images reach the eyes of human scientists.

This in turn helps to save the scientist days of manual photo review.

As the Myriad 2 chip was not originally designed for orbit, much work had to go into testing and preparing it for spaceflight. This involved a half-dozen different organizations across Europe, including the European Space Agency as well as start-ups and private companies.

But the end results appear to be worth it.

Spacecraft computers can be up to two decades behind state-of-the-art commercial technology due to their highly specialized hardware, but the Myriad 2 passed all stringent tests off-the-shelf with no modification needed. This includes 36 straight hours of blasting a radiation beam at the chip.

It was also not particularly difficult to train the AI algorithm, as the team could accomplish it based on synthetic data extracted from existing missions.

Thanks to Myriad 2, all this system and software integration and testing were accomplished in just four months. And in just three weeks after it launched, the team successfully verified that its implanted brain was behaving exactly as expected.

In the future, hardware-accelerated AI built on-board such tiny satellites could potentially be used to spot wildfires and notify local responders in minutes rather than hours. Over oceans, which are typically ignored, a satellite can spot rogue ships or environmental accidents. Over forests and farms, a satellite can track soil moisture and the growth of crops. Over ice, it can track thickness and melting ponds to help monitor climate change.

Many of these possibilities will soon be tested: a PhiSat-2, which will carry another Myriad 2 into orbit, is already under development. PhiSat-2 will be capable of running AI apps that can be developed, easily installed, validated and operated on the spacecraft during their flight using a simple user interface.

Read more about the PhiSat-1 on Intel’s blog here.