Video Above: Army Research Lab Advances AI to Land Drones on Tanks
A drone spots enemy armored vehicles approaching from other side of a ridge while functioning as an “aerial” node connecting otherwise non-line-of-sight combat assets, targeting data is then sent to an armed Abrams tank in position to lay down suppressive fire and instantly cue small groups of mobile, dismounted scout units in position to approach the enemy formation. The forward scout unit is then able to “paint” the target with a laser spot enabling nearby Air Force fighter jets to fire precision air-to-ground missiles to destroy the enemy positions.
“I may receive information from an intelligence system or sensor system, that's saying, Hey, I see a target on the ground that I want to shoot at. When I inform, I use a standard message. The problem with our community today is that sometimes the information that's in that message isn't good enough to fire on,” Alan Hansen, Deputy Assistant Director for Information Dominance, U.S. Army DEVCOM C5ISR, told Warrior in an interview
Joint All Domain Command and Control program (JADC2)
This kind of cross domain synergy, or high-speed information flow, is the conceptual foundation of the Pentagon’s fast evolving Joint All Domain Command and Control program (JADC2), yet its successful execution involves a key mixture of interwoven variables to include “gateway” technologies to “translate” or connect sensor data from otherwise incompatible transport layer communications technologies. At the fundamental technical level, interoperability or the “pooling” and organization of different streams of incoming data takes place at the level of “1s” and “0s,” meaning computerized standards and IP protocol engineered to enable common format information exchange. Much of this data organization and analysis process, which represents the core technological foundation of the Pentagon’s now “breakthrough” shortening of sensor-to-shooter timelines across the military services, is increasingly AI-enabled and driven by ultra-high speed computer processing.
Data Fusion - Project Rainmaker
Army DEVCOM scientists are improving upon, refining and advancing what they identify as a critical “data layer” or “Data Fusion.” This Data Fusion, referred to by Army scientists as Project Rainmaker, integrates otherwise disparate sources of information which are “collectively” interpreted and analyzed in relation to one another.
“When I say Data Fabric, I talk about a data layer that sits between the information systems in computing and our networks. That gives us the ability to come to define how and what we want to exchange information between,” Hansen said. “We're looking at trying to create an environment of analytics and Data Fabric, because you may have two different systems sensing something that's out there. The goal of a Data Fabric is to share the information between those systems.”
Perhaps older electronics or an RF signal, receiver or communications technology might initially seem incompatible with new technologies bursting into the operational scene? How can systems be upgraded and how can legacy or existing systems be successfully integrated with newer, faster, AI-enabled data analysis and transmission technologies? This is a key function of Project Rainmaker.
Ultimately, the process includes high-speed “collective” analytics wherein multiple separate variables and information “pools” are integrated into a single, clear picture for decision makers. This is the goal of JADC2 in the near term and also identifies DEVCOM research efforts aimed at expanding, hardening and accelerating this process for future war. Perhaps new transport layer technologies will emerge, perhaps data from previously impossible or unprecedented distances can be shared and analyzed across multiple domains. A key area of focus refers to efforts to enable the analysis to be performed at both central “hubs” and also at the “point of collection.”
“Each system will actually correlate the information and in the environment to take those correlated pieces, bring them back and actually adjudicate them into a single observation. Where does that happen? It can happen anywhere it can happen on the aircraft or on the ground, depending on where you deploy that analytic…so the analytics does the processing,” Hansen said.
The need to identify targets, communicate the details and “attack” or destroy could be described as a somewhat “timeless” reality of war, yet current technology is now bringing this process to new, breakthrough levels of speed and efficiency. The advent of AI, high-speed digital computer processing and multiple “transport layers” or sources of interoperable transmission technologies have made it possible to find and eliminate targets across vast, previously impossible distances, domains and areas of operation.
“With the Data Fabric, I can augment or include information into that topic that gives soldiers what they need in order to effectively target something…..so it sets the parameters. You can be very agile, what the parameters are,” Hansen said.
There has in recent years been an evolving and critical “Concept of Operation” or tactical dimension to this, as information management, processing and transmission has itself taken on new urgency as a weapon itself. Improving this process and engineering the technological foundation for ongoing modernization is the focus of the Army effort “Rainmaker.”
“Information dominance information today is a force multiplier. I think our leadership is starting to see that. The Army sees weapons, a tank gun or an aircraft as a weapon system. But data right now is becoming one of those commodities that is actually a system…and data is only good if you can serve it up right, to the right proponent at the right time, in order to make critical decisions,” Alan Hansen, Deputy Assistant Director for Information Dominance, U.S. Army DEVCOM C5ISR, told Warrior in an interview.
Army scientists and experts are looking at both near-term and longer-term breakthrough applications of data processing and transmission, a sensor-to-shooter time-shortening ability increasingly able to take place simultaneously among air, land, sea and even space domains. What used to take attackers as much as 20 mins, can now take 20 seconds or less, enabling offensive forces to operate faster or inside the decision-making cycle of an enemy to improve lethality and prevail in war. This goal or integrated, multi-service effort is referred to by the Pentagon as Joint All Domain Command and Control (JADC2).
“With JADC2, the intent is that we're going to fight as one service, not as the individual service. So weapon systems that have been managed by the Air Force might get data from the Army. Or the Army might receive data from the Air Force out of an aircraft and F 35,” Hansen said.
Recommended for You
Once quickly identified by AI-enabled systems bouncing streams of incoming new information off of a seemingly limitless data base to draw comparisons, assess possibilities in relation to historical precedent and ultimately recommend or “transmit” target specifics and an optimal “effector” or weapon with which to strike. The processing of information in a “collective” sense combining multiple sources of sensor data, once complete, can be transmitted widely across the force in a dispersed, yet networked and distributed manner.
“It could be an individual who's a squad leader on the battlefield, it could be a commander sitting inside of a TOC, directing a battle within their AOR. Or it could be a general who's actually managing the theater data. Part of it is becoming a very recognized force multiplier for us in future,” Hansen said.
An RF signal emitting from a soldier-handheld software programmable radio, a GPS signal traveling from space, point to point UHF datalink, optical transmissions or wireless internet connectivity can all deliver time-sensitive combat target specifics from different positions across different transport layer technological modes of transmission.
What if a unit confronting enemy fire in mechanized war simultaneously receives critical data from each of these disparate transmission modes? What if each source includes elements of an overall combat scenario picture needed by battlefield commanders and decision makers? What if an older antenna, radar system or computer cannot process data traveling on a different frequency or using a different data format? This problem has presented a fundamental quandary for weapons developers for decades, however the Pentagon and military services are now achieving previously unprecedented breakthroughs using common IP protocol and technical standards to, in effect, gather, translate, pool, organize, analyze and transmit otherwise disparate yet interwoven sources of combat-critical sensor data, fire control systems, weapons and large platforms.
Through Project Rainmaker, Army scientists are not only preparing for near term multi-domain interoperability through the use of high-tech, computer-enabled “gateways” or “translators,” but also working on new technologies intended to sustain breakthrough modernization for decades into the future. How will information need to travel between new, yet-to-exist platforms and transport layer technologies? How will it continue to achieve breakthrough speeds, distances and transmission capabilities?
“The whole intent of Rainmaker was built on the idea of sharing information at the very lowest level, to help build on this infrastructure of machine learning as we go forward. So it can accommodate as new stuff arrives, right, as opposed to being more narrowly restricted,” Alan Hansen, Deputy Assistant Director for Information Dominance, U.S. Army DEVCOM C5ISR, told Warrior in an interview.
Data Fusion, Hansen explained, is intended to help create a “collective” picture to an extent, and enable previously unattainable methods of data analysis and transmission.
“Using Data Fusion, you take different pieces of data and you correlate it, you combine the information into a single observation. Depending on what type of data that is, you may get a better observation based on different types of sensors for a different look at it from a different angle,” he added. “With a Data Fabric, I break it down into small pieces. And I may create a new methodology of decision making that we haven't ever done before.”
Hansen explained it in terms of an “interpretive” data layer to a degree able to combine and interconnect data from different incoming sensors traveling through different transport layers. It can also be tailorable, Hansen said, meaning it can be scaled to adjust to specific information-sharing needs and standards.
“Having a Data Fabric gives you the ability to kind of reinvent the way you want to exchange information. What type of information am I sharing? to whom, right? So these Data Fabrics give you this agility,” he explained.
An existing Army armored vehicle today typically has many “boxes” for sensing, receiving, fire control, computing and data transmissions, yet they often perform disparate functions using different hardware configurations. This can become problematic, as time-sensitive combat data can be segregated or “cut-off” between different “boxes,” “sensors,” or “methods of analysis.” This means vehicle crews and combat decision makers can get blasted or overwhelmed with unmanageable streams of data critical relevance to the fight.
In recent years, the Army has vigorously worked to address this predicament and lessen the “cognitive burden” places upon warrior-decision makers, freeing them up to focus on pressing variables in need of human attention. This involves lessening a “hardware” footprint and streamlining or analyzing different sets of information in relation to one another. DEVCOM scientists refer to this as “boxology,” a term referring to efforts to combine information gathering and processing “boxes” on a vehicle or platform. This can lower weight to improve speed and agility, while also opening up new, breakthrough avenues for information analysis and transmission. This process, referred to by Army leaders as Data Fabric, is part of a future-oriented technological Army modernization program called Rainmaker. A new “boxology” can enable unprecedented levels of operational functionality.
“Our boxology is too hard to break apart. So now we'll break it down into more puzzle pieces that we can put together and fit them as such. So our soldiers in the future can come up with new methods of how they want to use information in a way they can, how they do decision support, how they do fires, how they do logistics, for instance, when a Data Fabric, you can be a fires system, shooting off artillery,” Alan Hansen, Deputy Assistant Director for Information Dominance, U.S. Army DEVCOM C5ISR, told Warrior in an interview.
The success of the effort relates to messaging standards and formats, meaning the use of common analytical tools and gateway technologies to interpret, pool and analyze otherwise incompatible streams of information.
“The way we communicate today, with Boxologies with messaging standards, means that if I have to transform it, if I have to share information to another box, I have to transform my data into a message standard. And I move that metric standard across the network to another box, and it takes the data and it receives it,” Hansen said.
With data fabric and Rainmaker, however, different boxes, combat functions and information-gathering sensors can be integrated, combined and processed simultaneously in relation to one another in a smaller “box” or form factor.
Kris Osborn is the President of Warrior Maven - Center for Military Modernization and the Defense Editor for the National Interest. Osborn previously served at the Pentagon as a Highly Qualified Expert with the Office of the Assistant Secretary of the Army—Acquisition, Logistics & Technology. Osborn has also worked as an anchor and on-air military specialist at national TV networks. He has appeared as a guest military expert on Fox News, MSNBC, The Military Channel, and The History Channel. He also has a Masters Degree in Comparative Literature from Columbia University.