Human-Machine Teaming for Defense

Imagine this scenario: You’re the leader of a dismounted infantry team in combat on the front lines, and you’ve spotted a building on high ground. It looks abandoned—but is it? Instead of risking your team, you send in a robot. Using sensors and cameras, the robot feeds back data as it travels, confirming each room is empty. At the top of the stairs, however, it detects a suspected improvised explosive device (IED). You signal your team to avoid the building while the robot establishes a communications link to friendly forces nearby so you can call forward an explosive ordnance disposal (EOD) team to disarm the IED.

It’s a scene that the Department of Defense (DOD) is determined to bring to the front lines. “The DOD is purchasing thousands of military robots and other intelligent machines for use in conflict areas,” says Vice President Joel Dillon, head of Digital Warrior Solutions at Booz Allen. “We’re working on the toughest challenge—to integrate them into the ecosystem and equip them to be trusted teammates.” 

It will take diverse technical solutions to accomplish this, but every innovation has one requirement in common: an open and connected architecture that can integrate agile technologies and connect to a multidomain network. This is a key enabler to the concept of Joint All-Domain Command and Control (JADC2), the DOD’s initiative to accelerate decision making across services and domains, based on instantaneous sharing of data—enabling widespread coordinated robotic capabilities for greater power at the edge. 

Teams at the front lines are at the highest risk in battle. Approximately 90% of casualties occur within line of sight to the enemy, according to the Close Combat Lethality Task Force. Robotic teammates and smart machines offer powerful capabilities that can dramatically increase survivability while helping warfighters achieve the mission. 

Today’s machines range from mobile computing devices and unmanned aerial vehicles (UAV) to autonomous ground robots—for example, roving Packbots which assess combat and disaster sites. Intelligent machines can become force multipliers if equipped with multiple sensors, artificial intelligence (AI)-enabled analytics, and controls allowing for intuitive human teaming. That’s where robotic development is heading.

DOD is accelerating capabilities now to prepare warfighters to dominate in a digital world. The goal is to equip robotics, computers, and other devices to: 

• Strengthen information superiority (example: provide aerial photos of an adversary’s position)
• Reduce cognitive load (example: identify new targets in a firefight)
• Take on physical loads (example: carry a squad’s heaviest gear)
• Handle dangerous tasks (example: deliver supplies or conduct medical evacuation in a danger zone)

Once created, an open-architecture solution can be applied across diverse platforms and enable multiple functions.   

“It’s more than a matter of adding capabilities—it’s integrating them,” says Joel. “Those are some of the most interesting problems. For example, a UAV with a daylight imager equipped with analytics can identify an enemy combat vehicle and say, ‘That’s a target.’ If the drone also has a thermal sensor, it can tell the vehicle’s engine is running and conclude, ‘That’s not just a target, it’s also an active threat.’”

Beyond the promise of equipping individual assets, there’s the big picture of integrating capabilities across the services. To get to that multidomain future, the military must solve multiple problems. Here are some of the biggest challenges:

• Connections—combining multidomain sensors into the single JADC2 network
• Command and control—having the ability to control multiple systems to integrate combat formations
• Siloed systems—integrating equipment, sensors, and software for flawless warfighter control
• Human factors—replacing technical controls requiring a trained operator with intuitive interfaces
• Edge computing—ensuring connectivity in high-risk environments

Knowing that the future of war is data driven and information centric, we’ve created a multidisciplinary group working on solutions to address challenges across the digital battlespace. This gives us a holistic perspective to work on advances from secure mobile networking to human-machine relationships. 

“The ‘human’ part of human-machine is critical,” says Randy Yamada, a lead scientist in our robotic lab. “We’re a firm that cares about people, so we approach technology from the user’s viewpoint. That allows us to create innovations that work in the real world.”

He explains the human-focused perspective plays out in varied ways—for example, we create warfighter readiness solutions that take into account a trainee’s cognitive and physical health in addition to strategies and skillsets. And we’re out in the field helping test solutions and prototypes for multiple projects, getting feedback on what changes will strengthen warfighter trust.

Including veterans on our teams helps keep human factors top-of-mind. For example, as a former Army commander, Joel remembers what it’s like to lug around heavy equipment. And our developers are wary of cognitive overload: “How can we integrate sensors that supply critical information without distracting the warfighter from the mission?” says Randy. “And at a higher level, how can we overcome risks to ensure ethical AI? We’re working to resolve these problems.”

In addition to integrating internal teams, we’re continually growing our network of partnerships. Our technologists and tech scouting teams give DOD access to leading innovators from Silicon Valley startups to traditional government contractors. And for cloud, we provide secure solutions on leading platforms from providers like Amazon Web Services and Google. 

“A startup doesn’t have the roboticists, data scientists, cyber specialists, engineers, and experts in government mission and certification that we do. So we can help them connect with DOD to show their ideas,” says Patrick Ward, a software engineer. “And we work to create open-source connections and application programming interfaces that the government owns so agencies can upgrade quickly and cost-effectively.”

Some DOD agencies are furthering innovation in human-machine interaction, while others are still formulating their approach. From our work with a spectrum of defense organizations, we have discovered these principles help organizations move ahead:

• Connect with providers committed to open-architecture, government-owned solutions which keep you in control.
• When migrating to the cloud to enable future robotic integration, focus on cloud-native solutions to enable faster capabilities and built-in security.
• Considering robotic development? Look for modular solutions to get the highest value over time.