Quantum machine learning enhanced sensor combination for Earth Observation (QMLS-EO)

The Quantum Earth OBServatory (QEOBS) project will use a high-altitude balloon flight to test a 2U CubeSat designed to demonstrate how onboard data processing and machine learning can result in reduced downlink requirements. Using an array of sensors, the test will evaluate quantum and classical machine learning approaches for Earth observation tasks, including atmospheric gravity wave measurements and multispectral image classification and segmentation. The project aims to use a 30-qubit onboard quantum simulator. The ultimate goal is to fly the 2U CubeSat in low-Earth orbit. 

Problem Statement Reducing the amount of raw data that needs to be transferred between a spacecraft and the ground has many benefits (e.g., enabling autonomous operations, increasing speed of information transfer). The use of machine learning offers promise in this domain by allowing data processing to happen at the edge of the network (e.g., on the spacecraft), thereby limiting the need to download the data for ground-based processing. 

Technology Maturation The aerospace industry is just beginning to apply traditional machine learning models for mission-critical elements. This project aims to test the applicability and the benefits of such onboard, in-flight technology and improve upon it via quantum simulation/computing, which could benefit the entire aerospace industry. The flight test aims to raise the technology readiness level (TRL) to 5 or 6. 

Summary of Flight Test
2022-07-28 To our knowledge this was a world-first technology demonstration of a suborbital flight utilizing quantum machine learning on the edge by utilizing both quantum simulators onboard and in flight as well as IBM quantum computers on the ground via remote connection for earth observation and space research tasks. Our successful flight shows that ML and QML algorithms can indeed be used to detect and classify objects from low earth orbit on remote edge platforms. Furthermore, we demonstrated the remote execution and processing of earth observation data while only downloading the results to the ground therefore only requiring a low communications bandwidth.
2023-05-24 The Quantum Earth Observatory (QEOBS) mission was able to complete another successful technology demonstration flight where dams were detected by machine learning algorithm sand confirmed by quantum machine learning algorithms onboard. Additionally, we were able to train new quantum machine learning models in flight with real quantum computers on the ground. QEOBS was also able to detect atmospheric gravity waves which were also confirmed via the respective gravity wave quantum machine learning models onboard during the cloudy days of the mission. We also collected a large number of multi-spectral (RGB+NIR) earth observation images, which can be used for future projects.
 

This project aims to test the capabilities of a quantum machine-learning-enhanced sensor combination for Earth observation (QMLS-EO) versus traditional and hybrid machine-learning applications for selected Earth observation tasks. Onboard processing has the potential to greatly reduce the amount of data that needs to be transmitted to the ground. The project aims to use a 30-qubit on-board quantum simulation. This has the potential to benefit NASA missions, the commercial space industry, other government agencies, and the nation. 

Future Customers 
This technology has a potentially widespread end-user landscape: 
- Public- and private-sector Earth observation 
- NASA, other space agencies, and research institutions (e.g., U.S. Air Force Research Lab)
- National security satellite applications (e.g., U.S. Air Force)

This is early/mid-stage (TRL 5) — the most realistic path in is NASA SBIR/STTR, which funds small businesses and research institutions to develop technology aligned with NASA's needs (equity-free, phased funding). Check whether a current SBIR/STTR solicitation topic overlaps with this project's technology area, or contact the project directly (above) to ask.

None of these are guaranteed paths for this specific project — TechPort itself doesn't have an "apply" button. Reaching out to the contact(s) above with a specific question is usually the fastest way to find out what's actually open.