In everyday life, advanced technologies — such as machine learning and artificial intelligence (AI) — are so common that they fade into the background. Even though these advancements are fairly recent, they became ubiquitous so quickly that it’s easy to take them for granted.
In the battlespace, it’s an entirely different world. Much of today’s cutting-edge technology started as military research and development projects, and then evolved to become the gadgets and services we all know too well. But those consumer conveniences aren’t built for the battlefield. If your smart speaker misunderstands your command, it might play a song you didn’t want to hear. If that type of flaw occurred on a fighter jet or a tank, the consequences could be deadly.
This isn’t overstating the challenge. In the defense industry, security and reliability are a matter of life and death. At the same time, if adversaries are adopting the latest technology, militaries must keep up, or suffer a disadvantage on the battlefield.
Built on a Foundation of AI
Machine learning is a type of AI involving algorithms that improve on their own based on data and experience. The more practice they get, the smarter they become. These algorithms are the foundation for many of our favorite modern conveniences: personalized suggestions for your next favorite TV show, ride-service apps, self-driving vehicles and much more.
In the defense industry, AI could give troops more situational awareness and the ability to make faster decisions. The more the military can outsource to machines, the easier it will be for warfighters to focus on the big picture and the details that matter most for surviving and protecting their nation.
In some ways, AI is already in the defense toolkit (e.g., automated target recognition). Now, the next generation of AI for defense will upgrade even more tools, so machines can perform tasks that previously required human intervention.
Advanced Sensors and Connectivity
As the U.S. Department of Defense (DOD) moves toward making more advanced battle tools, sensors and connectivity will be two big pieces of the puzzle.
The DOD is looking to connect and integrate sensors with shooters and platforms across all domains, commands and services. The initiative, called JADC2 (Joint All-Doman Command and Control), is about adding sensors and connectivity to the battlefield and connecting those new capabilities with different branches of the U.S. military.
A big challenge here is adding sensors to legacy systems that are often intended to last decades. While consumers are used to buying a new phone every two or three years, militaries often operates multimillion dollar vehicles for several decades. And the sensors aren’t your everyday electronic components, either — they must have built-in security measures. In this case, rather than implement a sweeping update, sensors are gradually introduced to existing systems.
Plus, these sensors are only useful if troops can access the information they hold. A truly connected battlefield needs reliable, fast data transfer. You won’t find many 5G connections or Wi-Fi hot spots in the contested and remote areas where militaries often needs them.
An even greater challenge is security, because the more connected devices you have, the broader the attack surface you must deal with. More devices available to manipulate means more opportunities for hackers or bad actors to break into the systems.
That said, there are solutions to some of these challenges, and more are being developed today. For example, Northrop Grumman’s gateway systems are like “Wi-Fi in the sky.” They provide secure connections via airborne and tanker platforms, which can help mitigate risk.
Seeing Through the Weeds
As sensors and connectivity provide unprecedented amounts of information, machine learning helps bring it all together. In stressful situations, too much information is the opposite of helpful. Machine learning algorithms can help weed out the extraneous data to avoid information overload. As more technology is rolled out to the military, there is a delicate balance between connectivity, efficiency and security.
New solutions, such as Northrop Grumman’s DA/RC software, could provide decision-making for manned and unmanned fleets. For example, a leading prototype JADC2 system can connect and control a range of systems across domains and services.
This type of machine-learning capability, layered on top of trusted legacy systems, is shaping tomorrow’s tech-driven warfare. Machines could, for instance, observe the surroundings and decide on the optimal weapon for a scenario.
A Secure and Ethical Approach
To build these new systems and support, Northrop Grumman is using a DevSecOps approach in which development, security and operations are included every step of the way. The company is also collaborating with universities, such as Carnegie Mellon, for research and development, as well as startups, such as Credo AI, for cutting-edge technologies.
Underlying these various technologies, the government and industry partners are aiming for a secure and ethical rollout of artificial intelligence and high-tech tools. That means thinking deeply and strategically about letting machines take the wheel for tedious tasks while always keeping a human in the loop for critical decisions.
Are you interested in all things related to technology? We are, too. Check out Northrop Grumman career opportunities to see how you can participate in this fascinating time of discovery.