WASHINGTON — Scientific Systems, a defense contractor based in Woburn, Massachusetts, won a $1.5 million contract from the U.S. Space Force to develop software for in-space object detection and identification.
The company will work on the project with Stanford University’s Space Rendezvous Laboratory over the next 15 months, Scientific Systems’ vice president Owen Brown told SpaceNews Oct. 17.
“We will develop software that will be used in small cameras to allow satellites to safely identify, approach, and service other satellites or debris objects,” said Brown.
The contract is a Small Business Technology Transfer Phase 2 award from the Space Force’s Orbital Prime program, a research and development project focused on debris removal and in-space services technologies. Scientific Systems won one of 20 Phase 2 contracts awarded to date by Orbital Prime.
Scientific Systems is working on other military space programs, including as a subcontractor to SAIC in the Space Development Agency’s software app factory. It also developed a payload for an SDA demonstration of autonomous satellite operations that launched in 2021 on a Loft Orbital satellite.
Machine learning technology
For the Orbital Prime project, Scientific Systems is using a computer vision tool designed for autonomous inspection, tracking and identification of unknown objects. Stanford’s Space Rendezvous Lab is contributing machine learning technology.
Brown said the company’s long-term goal is to install the software on a space camera and test it in rendezvous and proximity operations (RPO) in orbit.
“This award puts us on a path to create RPO-in-a-box,” he said. That would be an integrated hardware and software product for spacecraft performing servicing missions, for example. Scientific Systems’ tool would help ensure that the vehicle approaches a client satellite safely, said Brown.
He said the technology has been demonstrated in the air domain with a sense-and-avoid software tool used by unmanned aircraft for safety of flight.
Simone D’Amico, founder and director of the Space Rendezvous Laboratory at Stanford University, said the project will help “create new capabilities for the rapidly developing on-orbit servicing and manufacturing ecosystem.”
“We will be integrating our machine-learning based relative navigation approaches with unique algorithms being created by Scientific Systems,” D’Amico said.
Sean Phillips, a technical adviser at the Air Force Research Laboratory, said the solution proposed by Scientific Systems and Stanford will help advance “next-generation autonomous satellite operations for close-proximity interactions like satellite inspection.”
