Turning Data into a Weapon Through Advanced Battle Plans

COTS Journal?

It has become clear that the future American warfighter will require advanced Artificial Intelligence or AI that stretches from military headquarters to the raging battlefield. This will be required in a seamless, adaptable, and ruggedized form.?As the defenses become more connected, a highly distributed computing ability at the edge becomes necessary.?Beyond systems that are adapted to extreme environments, however, the engineers at SECO have identified and are focused on providing three key capabilities that are needed.?These are complex, application-specific algorithms, curated data, and computing infrastructure.

Curated data refers to data that has been carefully gathered, sifted, chosen, and organized for its use in military decision-making. It must be so organized and formatted that it can readily be used by all present and future algorithms in the system. Algorithms must of course, also be written so that they can readily work with the curated data. The computing infrastructure must be rugged and readily configurable from the command center to a single soldier's sensor on the edge of the battle.

The engineers at SECO have identified six technologies that are relevant in addressing these stated goals:?

1.????Natural Language Processing or NLP performs speech and text recognition to extract relevant intelligence on troop position and other key data from enemy communications.?

2.????Computer Vision identifies and classifies threats accordingly.?

3.????Scheduling and Resource allocation optimize the response.

4.????Rule-based solutions create a large library of "if this, then that" solutions.

5.????Machine Learning acquires knowledge from curated examples and training.

6.????Robotics allows an embedded system the ability to interact with the environment.?

The Department of Defense has been riddled with false starts and countless acronyms in pursuit of a connected battlefield. The latest acronym that has been built on its predecessors is Joint All-Domain Command and Control (JADC2).?The important part is the C2 command and control.?The Joint all Domain means covering all branches of service and their assets on the battlefield.?Simply put, they want to have all assets/sensors talking to each other and a solution to be employed based on a compendium of data – or in other words, artificial intelligence.?

In recent years the predictability of how certain actions will be responded to has left us vulnerable as certain responses are baked into the strategy.?The enemy in knowing our playbook, may be anticipating our response to any aggression.?Employing AI will bring about situational awareness that provides an asymmetrical response leaving the adversaries surprised.??Not knowing a programmed response and thus being unable to choose an effective reaction quickly will cause the enemy to pause as their sacrifice will be unknown. Time can be a real killer.

Based on a long experience of solving discrete applications with rugged compute solutions, SECO quickly recognized the large data sets and the complex theaters of operation would inevitably lead to project creep.?With each additional connection, this potential grows exponentially, inevitably leading to a crippling complexity. The SECO approach uses industry standards as an array of building blocks that can be connected for a unified solution.?This enables a single element, such as a Hummer or Joint Light Tactical Vehicle (JLTV) to act independently and contribute sensor data up the command structure.?

The vastness of a connected battlefield goes beyond land, sea, air, and space to include command structures and policies.?It also includes cyber security threats that can do everything from spoofing a sensor to crippling whole networks.?Building from their experience, SECO uses open standards from a portfolio of off-the-shelf software and hardware solutions to allow them to address varied applications.???An early example of a SECO success is the DCD or deck control device, used to control the MQ-25 Stingray ? unmanned aerial refueler.??The user-friendly control unit exemplified an application-specific need that connected its user while meeting the strict technical requirements set forth by Boeing.?

The defense world is fraught with numerous unforeseen issues that range from traditional systems that do lend themselves to being connected to political mandates that prohibit certain branches of the service from reacting to certain kinds of incidents.?An example of the first is the reliance on verbal communication to control the airspace.?Transitioning from this to an AI-based solution would require a complete overall of the rules and policies that have been used for decades.?The narrow charter between services reflects another example.?Do you deploy a Navy ship, a Coast Guard cutter, or a Harbor Patrol boat from the nearest municipality in pursuit of a threat of a boat entering American waters??These are two of the many potentially difficult hurdles to overcome in bringing the past into the present.

In developing connected solutions, engineers have come to recognize that it isn't enough to automate the past. Solutions need to be looked towards the future.?This is never more important than in the defense of the country.??Today's connected battlefield will have to draw upon open-source data sets from sources such as those available from social media.??In balancing the needs of the past and future, SECO has gathered more than hundreds of the leading AI minds in the Bay Area of California to create software solutions that tame the beast into more manageable chunks with complex algorithms and an interpretive User Interface.??This allows SECO to set the conditions in the theater for future operations, including access, intelligence, plans, training, and interoperability.

The SECO approach is that AI advancements will be more driven by the need for data interpretation to recognize and measure threats.?These include intelligent and autonomous vehicles ranging from the Abrams Tank to rotorcraft.?Tasks like intelligence analysis and predictive maintenance will be integral and done discretely, for example, predicting the likelihood of encountering an Improvised Explosive Device (IED) and determining its location.??The fear of some runaway automated nuclear launch sequence is not something that should be dismissed, but the primary focus of the US and China is on data interpretation and how it might contribute to a country’s defense.?

SECO set out to reduce the complexity at the edge by creating a simple, intelligent sensor for the cloud known as Clea.?Clea is an AI/IoT platform that easily connects edge electronic devices with the cloud and facilitates real-time device monitoring, analytics, infrastructure management, predictive maintenance, secure remote software updates, and more. ?The objective of Clea is to offer a computing solution at the edge that can endure the rigors of the environment and yet still offer a competent AI computing engine connected to the cloud.?This simple and elegant solution can reduce the raw data from the front to small actionable data sets that support a strategy.??The propensity of the warfighter to believe that their instincts are more dependable and, consequently, they might want to override the system are mitigated by the Clea.???This occurs by actionable data being conveyed in a digestible fashion to the field.

In training for the use of AI in a combat theater, SECO set out to build a foundation of solutions that build upon each other.???This approach has resulted in confidence on a local level and conveyed it up the command structure.?The idea of a hierarchal command structure, as used in industrial applications throughout SECO's history, is ideal for the highly distributed command structure of the military.??Whether it is the Sensor-to-Shooter initiative of the Army or the Advance Battle Management System of the Air Force, solutions built from the bottom up are more in tune with the functional needs.?

The interoperability of systems leads to greater complexity and thus, more possible intrusions or potential faults.?The SECO approach utilizes industry standards that have proven to be fault tolerant as building blocks toward a unified solution.??The hurdle of interoperability within a command center is more complex with each additional connection.?For this reason, SECO’s approach of putting the most advanced AI solutions at the edge will better ensure that latency and response times are reduced.??This combination of data analytics and redundancy aids in affirming that sensors have not been spoofed and are correctly reporting the data.??This enables compound queries to be executed more safely and securely by using processing at the edge and identifying irregularities with encryption keys and data sets.?

An example of how strategically important the idea of network sensor-to-shooter is to the Department of Defense is the next-generation long-range strike bomber, the B21 Raider.?Although its primary function is to be a long-range bomber, the Air Force also wanted the plane to act as an intelligence collection platform for battle management.?The first pictures of the plane were released to the public in Early December 2022.

The 21st Century warfighter will rely on solutions that draw upon knowledge gleaned from all available assets. This will require a speed that outstrips our adversary’s communication structure and produces an asymmetrical response that is surprising and unsuspected.?These solutions will be built upon discrete activities defined using off-the-shelf solutions in a way that builds confidence.?From a complex response that may seem non-sensical to the warfighter to the identification of security breaches, AI at the edge will be integral to a superior defense.??

?[TW1]What is a DCD?