Artificial Intelligence for the Army’s Big Six Modernisation Priorities

Brought to you in partnership with Systel

The US Army’s modernisation priorities will fuel a massive expansion in the data generated by soldiers, units and military platforms. It has become clear that artificial Intelligence (AI) will play a key role in processing this data – and exploiting the opportunities it provides.

The army has identified six broad areas of focus for its future development: long-range precision fires; next-generation combat vehicles (NGCV); future vertical lift; army network; air and missile defence; and soldier lethality.

As technology evolves across these priorities, there will be a significant increase in the amount of data available to individuals, units, and beyond, said Aneesh Kothari, Vice President for Marketing at Systel, Inc.

Kothari pointed to NGCV, which spans several manned and unmanned vehicle platforms. As such vehicles are upgraded, they are increasingly outfitted with full HD, high-resolution sensor suites and cameras, such as EO/IR systems, generating full motion video (FMV) data. 

Those sensors collect immense amounts of data constantly, something that will only expand in the coming years.

‘There is an enormous amount of critical information being presented to the operator,’ Kothari said. ‘This results in potential for overwhelming information overload on the vehicle crewman.’

Stryker crewmen with the 1st Squadron, 2nd Cavalry Regiment (2CR) fire an M1128 Mobile Gun System during a live-fire exercise in Poland. The Rapid Capabilities Office is seeking new technologies that apply artificial intelligence and machine learning to paint a picture of the electromagnetic spectrum. (Photo Credit: US Army)

This is where AI technology and techniques can be utilized to help solve mission-data-processing workloads. The technology can process all vast amounts of data, filter and compartmentalise it in near real-time, ‘taking that cognitive burden away from the warfighter’, Kothari explained.

It can then present clean data sets to soldiers, empowering them to make actionable decisions, from engaging an enemy to taking evasive manoeuvres.

This AI capability rests on a hardware backbone, which must ingest, process, exploit, and disseminate the raw data at real-time speeds. 

Kothari pointed to software being used on NGCV demonstrator vehicles that focuses on hostile fire detection and localisation (HFDL) and which runs on Systel’s Raven-Strike rugged mission computer.

The software detects incoming projectiles or heat signatures, quickly providing the operator with the data and information necessary to make an informed decision.

Raven-Strike™ offers complete sensor integration and data fusion, in a SWaP-optimised system.

View full size

‘It takes immense compute power, robust I/O capability, and high-bandwidth networking, all combined in a SWaP-2C optimized single LRU rugged system,’ Kothari said.

This is a key focus for Systel: delivering rugged computers designed to integrate multiple sensors for data fusion in a single line replaceable unit (LRU) system and with the computing power necessary to support the required software applications.

The company considers a wide range of factors, from examining the networking capability needed to present the data (both inside the vehicle and downstream) to assessing the optimum storage capability.

It also focuses on a number of practical areas, such as optimising the computing system to fit in the space available and ensuring it has the necessary thermal technology to avoid overheating.

‘That’s how we enable the AI/autonomous capability, through understanding the overall platform ecosystem and tying in the compute piece.’

AI has a range of other uses, Kothari noted, such as boosting cyber security and performing biometric analysis, for example facial recognition. The exponential multiplier effect of deployed AI empowers operators to turn reams of raw data to their advantage.

When it comes to NGCV and other platforms, these capabilities ‘must take place at the edge’, Kothari said, relying on an onboard computing system without a cloud storage server to provide feedback and support.

Commander's Virtual Staff will explore commercial technologies and advances in artificial intelligence as part of its efforts to support tactical decision making for US Army commanders and their staff. (Photo Credit: US Army)

That same capability and technology must be available in the vehicle itself, perhaps without strong connectivity.

‘It’s critical from a soldier protection perspective to not only have the compute capability necessary to accomplish the task at hand but also for that capability to reside in a fully rugged system designed to withstand the harshest environmental conditions,’ he added.

Systel’s major product line for such applications is its Strike series of rugged mission computers, featuring systems such as Raven-Strike and Hawk-Strike, designed for mission-critical applications featuring multiple sensors and network interfaces.

The Strike series is built with AI applications in mind, Kothari said, with enough processing capability ‘to help shift the focus from man to machine’.

Systel has supported AI technology in various other projects in the past, including Aurora Flight Science’s Autonomous Aerial Cargo Utility System (AACUS), a UH-1 helicopter capable of autonomous flight.

Its rugged servers have also been used in the Pentagon’s Project Maven and the US Navy’s MQ-8C Fire Scout unmanned autonomous helicopter. 

An M2A3 Bradley Fighting Vehicle crew changes position on the range during gunnery training. The US Army is developing a new Next Generation Combat Vehicle as part of a concerted modernization strategy. (Photo Credit: US Army)

AI is set to evolve in military use at a rapid pace, said Kothari. It will not only lift the cognitive workload from the warfighter, but can also remove them from front line danger altogether, something the NGCV Cross Function Team (CFT) is working to achieve with the Optionally Manned Fighting Vehicle (OMFV) and Robotic Combat Vehicle (RCV) future platforms.

Systel’s Raven-Strike computer is successfully hosting the HFDL system being tested on the OMFV Mission Enabling Technologies Demonstrator (MET-D) and RCV surrogate vehicles.

A number of complex technical challenges must be overcome to make this future a reality, Kothari said.

Just as self-driving cars must learn to recognise lines on the road even in snow or rainy conditions, so military vehicles must be capable of determining whether particular terrain – be it hills or sand – is safe to traverse.

‘All of that capability has to reside on a mission computer,’ he said. ‘That’s what we provide, and that’s how we support the armed forces.’