Military innovators are shrinking the sensor-to-shooter gap from minutes to seconds, leveraging edge computing to identify and neutralize fleeting battlefield threats before they vanish from sight.
..............“Very quickly make the best decision with the best information “........
By Kris Osborn, Warrior
(Washington, DC) If an enemy tank were to briefly emerge into an open area between two wooded areas for roughly 30 seconds before again disappearing from view, would a small forward aerial sensor, attack drone and fighter jet have enough time to close the sensor-to-shooter gap, verify the target, identify the optimal weapon … and destroy the enemy tank?
The answer is increasingly “yes” given the rapid arrival of AI-enabled target and data processing at the forward edge of combat. However, AI alone has little immediate tactical relevance without a transport layer “networking” system to “transmit” time-sensitive data fast enough and accurately enough to close the combat loop with success. Data can be collected and processed by AI-empowered sensors and small form factor, mobile servers and processors — yet it needs to “reach” the effector or accurate weapons system in time to destroy the target in a matter of seconds.
This is why Pentagon and industry innovators are moving quickly to better integrate highly successful and promising AI applications with new generations of Command and Control technologies and tactics.
“The old model was…. a sensor would pick something up, and that data would go back to the data center for some processing. Usually there's a human in the loop, as someone has to make sense of it. Then the shooter, either gets a call or receives the data and makes a decision to act. That takes several minutes at the very least. With the new model, you know, we're pushing that thinking to the edge, right? AI allows that thinking process to be at the edge where the sensor exists….So we’ve cut that time down dramatically. It's going to be flagged data received by the shooter in just a matter of seconds now, instead of minutes.” Kimberly Lowe, AI Architect,for Strategic Planning Office, Curtiss-Wright Defense Solutions, told Warrior in an interview.
This kind of AI-Command and Control synergy is a huge priority for U.S. Army Pacific’s new Multi-Domain Command - Pacific unit created earlier this year. Service leaders stress the reciprocal relationship between AI-enabled computing and analytics, and transport layer data transit and interoperability
“If we’re going to have a diverse sensor array and ensure those sensors are in the optimal locations based on signals of interest, AI is really helpful to get them into position without multiple analysts manually maneuvering those assets. If we’re tying that to agile effects formations—offensive and defensive—we need to absorb and make sense quickly of what those sensors are seeing from space, stratosphere, air, maritime, and terrestrial domains, and then process and rapidly push that to the best shooter based on location and magazine depth. AI is helping us do this,” Maj. Gen. Bernard J. Harrington, Commander of the 7th ID, Multi-Domain Command - Pacific, told Warrior in an interview.
AI & The Tactical Edge
An ability to manifest this forward processing and analytics would seem to require both hardware and software applications as incoming data from otherwise disparate transport layers and sensors needs to be aggregated, organized, analyzed and transmitted. This is one reason why the Pentagon and industry are rapidly bringing AI-enabled technology to the forward edge. This amounts to the construction and delivery of small-form-factor, even hand-held devices capable of operating as forward-positioned AI-enabled servers and processors. One of many industry efforts to harness and actualize forward AI processing, such as that described by Lowe and Alan “Doc” Docauer, Director of Growth, LYNTRIS Business Systems” is Curtiss-Wright’s emerging PacStar 810 server/processor AI-capable "box." In some instances, service members on the tactical edge will be empowered with the data they need to "strike," "defend" or "maneuver" in response to enemy activity themselves in a matter of seconds. In yet other instances, the PacStar 810 enables transmission of `organized, time-sensitive and highly relevant information to command control centers or integrated "hubs" to optimize the decision-making process. As a small, 20-pound transportable forward-operating "server" application with AI-capable computer processing.
“There are some tools that are emerging that are literally Gen 2 or Gen 2 Plus AI that are going to help sort through data instantly to get all that decision quality information. This gives the operator, the battle manager, at the front end of the kill chain an ability to make sense of the operating environment and get that decision quality information. Your systems are now helping you to understand and, and contextualize the decisions you have to make,” Alan “Doc” Docauer, Director of Growth, LYNTRIS Decision Systems, told Warrior.
This migration to the tactical edge and reduction of a hardware footprint with smaller form factors represents a key area of breakthrough innovation in recent years. At the U.S. Army’s “Project Convergence” in 2020, for example, sensor-to-shooter timeline breakthroughs were reduced from minutes to seconds using a central ground-based AI system known as FireStorm. This computer gathered incoming sensor data from drones, mini-drones, helicopters, armored vehicles, satellites and dismounted soldiers and performed the necessary analytics in near real time to “recommend” optimal shooters, weapons and effectors tailored to a particular threat circumstance.
In the six years since Project Convergence 2020, the Department of War and its defense industry partners have advanced paradigm-changing innovations which leverage the progress from 2020 and enhance it for smaller, more dispersed and more expeditionary forward maneuvers.
This changes concepts of operation and maneuver formations as more dispersed and “networked” AI-enabled analytics generates a faster and more streamlined information flow. Therefore, while AI is a key impetus driving the massive reduction in processing time and sensor-to-shooter pairing, the breakthrough progress “must” manifest through command and control networking and “transport layer” connectivity. Command and Control and AI need to be interdependent and intertwined for breakthrough tactical progress in combat.
“It's all underpinned by the need for resilient and assured networking…..so AI alone isn't the solution. It's a part of a number of solutions,” Docauer said.
Production & Innovation
When it comes to engineering the interfaces between AI-enabled servers and processors and command and control transport layer applications, industry partnerships are becoming increasingly vital. The idea is, quite simply, to ensure that open or interoperable interfaces and common IP Protocol standards are implemented to seamlessly “integrate” data across platforms and domains in support of a joint fight.
Partnerships are critical to “coordinate” and optimize battlefield “effect,” Lowe explained.
“The important thing to remember here, though, is none of the companies out there built the entire chain themselves…..partnerships are imperative……one company can build the server hardware, one company might do the networking, but it all comes together with the software. So those partnerships and those collaborations and integrations are key to making this all possible at the edge. None of us can do it by ourselves.” Lowe said. “A good example being we've got all of these ruggedized servers that we can create within defense solutions.They're useless without software right. But if we don't stick to the standard then the software developers don't know what they're dealing with either.”
While there has been much progress within the defense industry as companies migrate more fully to open systems architecture, there is much work to be done in the realm of ensuring continued, long-term interoperability and modernization, Doc argued.
“We have a lot of work to do both as an industry and to help the government to bring that to life. Open mission systems have the promise of allowing you to build software systems and do system level integration across domains, across systems and across the joint force with partner and allied networks. This can truly create a collaborative environment and create a place where data can be federated and shared into different software suites at different nodes,” Docauer said. “Industry has a lot of roles to play in that and includes you know companies being willing to kind of forego you know the change their mindset about how they sell stuff systems so we are going to non-proprietary open architecture or open systems.”
As AI-empowered systems continue to “surge” into production and impact combat and concepts of operation, now is a critical time to ensure more ubiquitous use of open architecture through the military services and their industry partners.
“As long as we stay with the standards and we design to the standards, then any upgrade that we make to our hardware right gets propagated with the same standard. Now is a really good opportunity for us as we're developing all of these new packages and all of these new servers for AI. We can bake in the standards to make sure that we just save ourselves a lot of pain and help the customer going forward,” Lowe said.
