Ball Aerospace and Microsoft Show That the Cloud Works in Space

Data from low Earth orbit (LEO) satellites in space can be processed, exploited and disseminated efficiently using a ground-based cloud computing system, according to recent studies performed by a team of Ball Aerospace and Microsoft researchers. The findings indicate that with the right technology, information gathered in space can be quickly and easily utilized to improve applications for a wide range of industries, from agriculture and sustainability to disaster response.

“Our tests showed that the cloud is ... not only fast, but also flexible, secure, scalable, and resilient,” said Steven Smith, vice president and general manager of systems engineering solutions for Ball Aerospace.

A Telesat teleport facility in Mount Jackson, Va. Data was delivered directly from space to an Azure Stack Edge device in this tactical vehicle. (Image credit: Microsoft.)

The technologies used to run the studies included Microsoft Azure, Microsoft’s cloud computing system, and Azure Stack Edge, a device that processes data and runs machine learning models at edge locations. Edge locations are those that are close to the end user.

The studies, conducted between February 2021 and May 2021, executed demonstrations by transmitting simulated overhead persistent infrared (OPIR) data via Telesat satellites to Microsoft Azure in a data center and directly to a tactical vehicle equipped with an Azure Stack Edge device. OPIR uses sensors in the infrared band to monitor missile launches and other military operations.

A rendering that shows how data from space could be sent through the cloud and then be processed at different locations. (Image credit: Ball Aerospace.)
A rendering that shows how Microsoft Azure and Azure Stack Edge interact with Ball Aerospace’s equipment. (Image credit: Ball Aerospace.)

After images were pushed to Microsoft Azure and the Azure Stack Edge device, Ball processed the data using event-driven architecture (EDA).

Rendering of an Azure modular data center in a remote setting. (Image credit: Microsoft.)

EDA is a software system that is built to capture and process events or changes in state. The EDA used machine learning algorithms to both process the data and detect certain activities and features. The algorithms generated insights that were then converted to messages.

The last step of the project involved employees of Telesat, the company that manufactured the satellites, joining the Ball-Microsoft team. Together, the group completed a direct downlink of data from the network of LEO satellites to a Ball-built electronically steerable Ka-Band phase array. A phase array is a computer-controlled array of antennas that can be electronically steered to point in different directions without moving the antennas.

The phased array was affixed to the tactical vehicle to show how actionable data could be delivered to service members based at different locations.

“Direct satellite-to-cloud communication and accelerated ground data processing allows the (U.S.) Department of Defense to gain advanced analytics capabilities. (These will enable) predictive modelling and new actionable insights capable of reshaping the future as they (DoD) advance their mission,” said Tom Keane, corporate vice president of Azure Global, which is part of Microsoft.

This project, which began in May 2020, is a defense-related collaboration to develop a prototype for edge devices for the Commercially Augmented Space Inter-Networked Operations (CASINO) Program Office. CASINO serves as the Space and Missile Systems Center (SMC) focal point for dispersed LEO technology and prototyping.

CASINO’s work is facilitated by the Defense Innovation Unit, a U.S. Department of Defense organization that develops commercial technology for the U.S. military. The goal was to show that cloud processing can be agile enough to meet the needs of the U.S. Space Force.

“For years, the military has envisioned an agile and connected force structure. During the demonstration, the CASINO team proved that we are ready to field low-latency links today. (That) moves this vision much closer to reality,” said Ball Aerospace’s Smith.

Steven Smith, vice president and general manager of systems engineering solutions at Ball Aerospace. (Image credit: Ball Aerospace.)

“Low latency” means a computer network is optimized to process many data messages with a low amount of latency, or time delay.

The Ball-Microsoft team said the technologies utilized present several advantages. They allow for a higher number of satellite streams to be added with little disruption to speed. This aids with the speedy analysis of different types of satellite images. In most of the test scenarios, data was captured, processed and redeployed five times faster than the customer’s established target speed.

The team further shared that the system technologies were built in a flexible and scalable way. This means that researchers can add or replace algorithms without significantly altering the system. Since the system was built in Azure and worked at the Azure edge, development time was reduced and the system was more resilient. The progress made by the Ball-Microsoft team provides the U.S. Department of Defense with additional capabilities to support and shape the future of the agency’s mission.

Cloud technology in space is becoming a significant industry trend. It requires the development of new technologies that include edge computing and satellite networking. To learn more about this trend, read: