WASHINGTON — The Defense Department’s Golden Dome program office is tackling the “real challenge” of “affordability” via acquisition reform, with a key focus on spurring development of artificial intelligence, according to a senior Pentagon official.
Marcia Holmes, DoD deputy director for Golden Dome, told the Miami Space Summit today that innovative approaches to acquisition are foundational to lowering the costs while quickly ramping up capabilities to integrate data from missile tracking satellites, develop space-based interceptors (SBIs) and automate a comprehensive command and control (C2) network to link sensors to shooters.
“The real challenge that we have today is how we integrate all of these space assets into an integrated architecture to provide a layered defense affordably and at scale,” she said. “To enable the acceleration and innovation of Golden Dome … the secretary has provided a very comprehensive and sweeping acquisition transformation strategy.”
AI and autonomy, meanwhile, are “going to play a larger role,” she said.
“It will change how we deploy and use our weapons. Increased autonomy will help us reduce our manning and in turn, will reduce our sustainment costs. Increased use of AI will allow us to parse large amounts of data and help to rapidly present options to decision-makers for the appropriate response,” she elaborated.
Echoing remarks made on Jan. 23 by Golden Dome czar Gen. Michael Guetlein at the Space System Command’s annual Space Industry Days in Los Angeles, Holmes said that the program office “acknowledges” that to “reduce the cost equation” DoD has to “pursue technology and innovation that will reduce our cost-per-kill and enable a layered, reliable, affordable defense in depth [and] weapon designs that can address multiple phases of threat trajectories.”
SBIs: A Key Cost Driver
Multiple outside analysts and lawmakers have raised concerns in particular about the cost of deploying enough SBIs on orbit to realistically have a chance of shooting down more than one or two enemy missiles launched simultaneously. The laws of physics dictate that interceptors in low Earth orbit (LEO) only stay over the horizon within shooting distance of a ground target for only about seven to 10 minutes. Thus multiple interceptors are required to target each adversary missile.
This problem, known as the “absentee ratio,” is particularly acute for space-based interception of adversary ballistic missiles in their boost-phase just after launch because the window for a hit is only a few minutes long at best.
Therefore, while the Space Force on Nov. 25 granted awards to 18 undisclosed contractors for boost-phase SBI prototypes, it followed up on Dec. 7 with a request for prototype proposals to industry for mid-course interceptors using kinetic energy kill vehicles.
Holmes stressed that AI will be particularly important for empowering a C2 network capable of relaying data in near-real time from missile tracking sensors to SBIs and other shooters on the land, sea and in the air.
“Another significant initiative for Golden Dome is the instantiation of an integrated, advanced and a resilient battle management command and control powered by AI, artificial intelligence, to coordinate the vast network of sensors and interceptors that will be part of Golden dome. The C2 is a crucial element to the success of Golden Dome,” she said.
Missile Tracking to Include Commercial Solutions
Holmes further noted that a “key element” for Golden Dome will be developing and integrating data from tracking satellites developed by both DoD and private firms.
“It includes leveraging and upgrading our existing systems while pursuing commercial solutions and building advanced capabilities for future threats,” she said.
Holmes did not go into detail, but the Space Development Agency (SDA) is developing a Tracking Layer constellation optimized for tracking both ballistic and hypersonic missiles in LEO, with the intention of deploying a network with global coverage beginning in 2029. SDA further is developing “fire control” sensors capable of providing highly precise targeting data to missile defense platforms.
However, its missile tracking effort has come under some criticism from the Government Accountability Office for underestimating the risks of tech development hurdles, schedule delays and cost overruns.
Space Systems Command also is working on a new set of some 30 tracking satellites with the same mission but stationed in medium Earth orbit able to provide persistent global coverage by the early 2030s. The command in June 2025 awarded $1.2 billion for a second generation of 10 satellites, called Epoch 2, of the service’s planned Resilient Missile Warning and Tracking (MWT) Medium Earth Orbit (MEO) program, to BAE Systems Space and Mission Systems. The birds are expected to begin launching in 2029.
Most recently, the Pentagon’s Defense Innovation Unit (DIU) on Wednesday announced that it is seeking “high fidelity” commercial sensors not just for tracking ballistic and hypersonic missiles, but also for discriminating between actual missiles and dummies as well as for fire control.
The effort has two key objects, the announcement said:
- Threat Detection, Tracking, and Discrimination: Demonstrate capabilities to detect, track and support the engagement of Intercontinental Ballistic Missiles (ICBMs) or Hypersonic Glide Vehicles (HGVs) across multiple flight phases (boost, midcourse, and glide). Advanced technologies to deliver the precise positional data and target characteristics required to reliably discriminate lethal payloads from non-threats, such as debris and countermeasures.
- Fire Control Enablement: Provide high-accuracy, real-time tracking data (including precise range, angular resolution and high update rates) necessary for successful Kinetic Kill Vehicle (KKV) engagements in endo-atmospheric and/or exo-atmospheric environments. Key characteristics for consideration:
The announcement stresses that the effort has an “aggressive” schedule, with prototypes ready for lab demonstration within six to eight months of award and on-orbit hosted payload demonstrations within 12 to 24 months. Interested vendors have until Feb. 17 to respond with proposals.
