Robotics and autonomous systems (RAS) and artificial intelligence (AI) are set to change the way the US military fights, from improving acquisition and delivery of mission critical intelligence to accelerating all-domain command-and-control and the fundamental aims of JADC2, enabling commanders to make faster, more informed decisions in battlespace.

As the US pushes forward to realise the full potential of emerging technologies, the AI revolution in military operations has already begun. For reasons both practical and ethical, mundane and dangerous tasks are increasingly falling to robots over human resource.

Uncrewed vehicles and aircraft are performing reconnaissance and providing an alternative to exposing humans to dangerous environments. Likewise, this simultaneously increases tactical advantage through technical surveillance, as footage captured of the surrounding environment and enemy positioning is fed back to human analysts.

Drone surveillance is gradually supplementing the role of the human scouts, providing benefits to military operations – in the first instance, it’s possible for machines to go where humans cannot, opening up the more dangerous and difficult-to-reach routes. In addition, use of machines reduces risk of human error, misinterpretation or inaccuracies in reporting and communication.

In fact, it’s the chain of reporting and communication that currently limits the potential of available technologies, and maintains the burden on commanders, who rely on accurate, real-time information to respond to threats quickly in the chaos of battlespace.

There is now a critical need to integrate AI and mission-specific algorithms to facilitate rapid front end analysis, to transform information into actionable intelligence to enable rapid response to threats and enable greater cooperation and information exchange between forces.

Commanders live by the OODA loop: observe, orient, decide, act. This means battlespace information or intelligence must be timely if it is to be actionable.

At its most fundamental level, AI is a vital asset to modern militaries for its ability to process vast tracts of information faster and more accurately than human analysts, relieving commanders of delays caused by cumbersome processes. In addition, AI can spot patterns and predict outcomes of potential courses of action.

A perfect example of AI’s ability to predict outcomes and pathways is evidenced in the US Air Force’s Advanced Battle Management System (ABMS). AI applications across the ABMS infrastructure enable commanders and warfighters to interpret the battlespace around them and to respond proactively as opposed to being on the back foot and reacting to situations too late.

For instance, the ABMS can identify and continue tracking a high-value target even when the target disappears from line of sight. By computing known variables, the AI can predict where a target will be located later. This enables commanders to mobilise assets well in advance and thereby steal or maintain the initiative.

The drawbacks of too much unfiltered information were highlighted in a recent InsideAIR podcast with Air Commodore Jez Holmes and Professor Patrick Baker from the RAF Rapid Capabilities Office.

A human is only able to deal with a certain amount of input before becoming overwhelmed. Too much information is, perhaps, just as paralysing to decision-making as not enough information.

But with sensor fusion and front-end tech, it’s possible to deliver ‘actionable intelligence’, rather than just large data streams for retrospective human analysis. In practice, this means machines will bear the burden of analysing large amounts of data and sensory input, while leaving only information that is strategically valuable and aids commanders in faster, more informed decision-making.

Additionally, further integration of AI and sensor fusion into ISR platforms supports the US military’s aspirations for JADC2 and all-domain command-and-control by providing complete environmental awareness and analysis instantaneously, integrating sensory readings in battlespace with intelligence received from the operational and logistical own forces, allied forces’ units and incumbent civilian agencies.

This, and a step further in front-end analysis capability, is also crucial to meet the aims of C5ISR systems and enable greater cross-domain cooperation. The capability of AI-enabled front-end analysis to transform large amounts of information into valuable, actionable intelligence will help to reduce the cognitive load upon the operator. This will be crucial to help commanders across all domains to direct forces, collaborate as necessary, and take the right action at the right time.

// Image credit: US Army