November 1, 2007  

Robots make war more survivable

No one can accurately predict what the robot-inhabited battlefield of the future will look like in five, 10 or 20 years. But we can look at the technical history of military equipment and make some educated guesses. Over the past two decades, two significant developments have advanced modern warfare. The boom in smart weapons has been a giant leap forward. But although they offer the ability to precisely target and other benefits, smart weapons also leave out something critical: the up-close and personal understanding of what’s on the other end of that weapons delivery. As a result, ironically, it’s the technology that was designed to keep war fighters out of the mix that is part of the reason there will always be human armies.

The proliferation of virtual presence technology has also had an enormous effect. From overhead sensors and unmanned ground systems to everything in between, virtual presence enables modern machines to listen, talk, see, com¬municate and network. These advanced communication capabilities will continue to evolve, and so will their effect on the way battles are fought.

Smart weapons and virtual presence technology are being driven in part by changes in the American public’s tolerance for casualties and prisoners of war, which is much lower than it has been in previous conflicts. In World War I, our soldiers charged out of trenches in the face of machine-gun fire with the hope that a sufficient number of them would survive to be able to assault the enemy on the other end of the battlefield. Fundamental changes in warfare ideology, strategy and tactics help explain why robots are being adopted for a range of mis¬sions; they keep war fighters at safe standoff distances. This substantial shift also provides insight into how robots will be used on the battlefield in the future.

Robots are already a critical security and intelligence com¬ponent of 21st-century warfare. On the battlefield and in war games, robots have proven that they make two important con¬tributions. Using a robot, you can much more quickly estab¬lish situational awareness. But what is surprising is a robot’s ability to increase the tactical speed of a mission.

Let’s use clearing a building as an example. If you did it the conventional way, you might use a squad of 12 guys. It’s a daunting and slow task because you’d have to cautiously make your way through the building without knowing what to expect. It’s an urban warfare task that has changed little in 100 years. With a robot, you can increase tactical speed, because if you search a room and there’s nothing there, you can quickly advance to the next room, knowing you can do so without opposition. You can continue advancing quickly through room after room after room until you find something. Once you find insurgents, an explosives cache or something else, then you know you have some work to do. But it’s that combination of fast situational awareness and increased tactical speed that make robots ideal for a range of missions.


The changing nature of warfare has had a profound effect on the adoption of unmanned ground vehicles (UGVs) and is indicative of the way robots will be developed and deployed. It’s interesting to look at the market maturation of unmanned mili¬tary vehicles, particularly the huge difference in the way soldiers have embraced UGVs compared to the way aviators originally rejected unmanned aerial vehicles (UAVs). When they were introduced about 30 years ago, unmanned air systems were shackled by the pilots’ union. I know that’s the case because I was one of those pilots back then. The differences in the types of missions performed by air forces and armies — and their frequency — are two factors that have heavily influ¬enced the differing adoption experiences of UAVs and UGVs. It’s not that aviation isn’t potentially dangerous; it is. But avia¬tion is sometimes described as hours of boredom interspersed with moments of stark terror. For the Army, certainly in places such as Iraq and Afghanistan, the moments of real danger are more closely packed. It’s more dangerous in a personal sense, plus it’s dirty, tiring and all those things that make the Army’s missions so challenging. As a result, soldiers have been much quicker to embrace UGVs as tools they can use to contribute to their survival.

In 2002, before the PackBot robot’s first deployment in Afghanistan, the tactic for doing cave reconnaissance was to tie a rope connected to a 12-foot stick around the waist of a 19-year-old soldier and send him in to see what was happening. It’s easy to see why the robot was an attractive tool for soldiers and why the Army was much quicker to embrace UGVs than aviators were to accept UAVs. It took more than 20 years for the market for UAVs to reach $500 million but only half as long for UGVs to surpass that same milestone. The markets for unmanned underwater vehicles and unmanned surface vehi¬cles are speeding toward the half-billion mark even more rapid¬ly. The adoption of UGVs illustrates what robot companies must do to transition from innovators and early adopters to the early majority — what’s called “crossing the chasm” in a book of the same name by Geoffrey Moore that talks about marketing and selling disruptive products such as robots to mainstream customers. To participate in defense acquisition, robot compa¬nies need an array of competencies, such as logistics, supplier management, and other infrastructure and management and design tools. Many robot companies with solid science and engineering capabilities and innovative culture still lack the ability to transform those assets into marketable products.

One bellwether for all military systems, including robots, is a digital architecture. The F/A-18 aircraft’s digital architecture, for example, permits it to carry more than 50 different payloads and perform many different missions. Likewise, a digital architecture gives a robot that same degree of adaptability, accommodating payloads, sensors and electronics that allow it to perform a vari¬ety of missions. Explosives-detection technology integration on a robot platform is one of the most recent breakthroughs.

Some significant but surmountable challenges are tackled that will help advance the development and deployment of military robots. Requirements and resources challenges are front and center right now; it’s easier for the military to contin¬ue funding what it’s already been doing than to fund a disrup¬tive technology. So change management is one major part of it. That change is happening, though, slowly but surely.

A second challenge also being overcome is the lack of peo¬ple skilled in robotics to meet the growing demands of the marketplace. Educational programs and degrees in robotics at MIT, Carnegie Mellon, Worcester Polytechnic Institute and other institutions of higher learning have finally started to alle¬viate that shortage.

Both of these challenges highlight the fact that the robot industry is young and still emerging. The parallels drawn between the adoption of computers 30 years ago and robots today by Rodney Brooks, iRobot’s co-founder and CTO and the former director of the MIT Computer Science and Artificial Intelligence Laboratory, provide a perspective on where we are and where we’re going.

Other situations also will affect military robots, resulting in noteworthy ramifications for robot companies, the industry and the military. Funding, of course, is a key driver of product and industry development. Major initiatives, such as the Army’s Future Combat Systems program, are funding the future. You’ve either won one of those contracts, or you’ve missed out.

A second interesting situation I foresee is a shakeout of the market similar to those in other industries in the past. Take the aviation industry as an example. In the 1930s, there was Lockheed, Martin, Boeing, Douglas, Curtiss, Wright and a bunch of other independent airplane companies. Over time, they have condensed into the smaller group of major aero¬space and defense suppliers. That same kind of compression will happen in military robotics, but much more quickly than it did in aviation, in part because the requirement to scurry across that chasm is now critically important. To survive the shakeout, it won’t be enough to be an innovative company — you’ll have to be a company that can produce and also demonstrate the competencies to produce for a major pro¬gram of record. Most of the original aviation companies I mentioned had never even heard of logistics at that time. However, to be a successful producer now, they all must have a competency in logistics.

The market shakeout will yield some benefits, including higher quality, more reliability and better-managed programs. It will also cause some significant merger and acquisition activity in the industry. A downside is that it’s going to become more difficult for other robot companies to successfully cross the chasm, because they won’t have the necessary competen¬cies. That’s going to mean that only a few of those innovative companies currently doing research and developing products are ever going to bring them to market. We have to think nationally about ways to funnel in those and other innovative ideas that will otherwise get excluded.


Nevertheless, the future holds tremendous potential for mili¬tary robots. The Explosive Ordnance Disposal market has been a great place for robots to prove themselves. But robots are available to only a small number of the potential users at this time. There is a much larger opportunity to use robots to pro¬tect more troops. Many engineers could use robots for route clearing and other missions they perform. Even larger is the potential for infantry to use robots as scouts or point men to gather tactical reconnaissance. The robot can go in first, estab¬lish situational awareness and let the troops know what they face so they have time to plan how to deal with it.

With the addition of more robots, the future battlefield will be more survivable. Conflicts will be fought with greater standoff distances. They will have better situational awareness and better coordination. We’ll still have soldiers on the battle¬field, but they’ll be able to leverage intelligent machines more so in the future than today — and we already do a lot today with satellite feeds, tactical reconnaissance, standoff weapons systems, airstrikes and more.

Those intelligent machines are sure to include autonomous robots. Now, all the deployed military robots are tele-operated, which means there’s a one-to-one relationship between an operator and the robot. We’re in the early days, but we are starting to build the autonomy. That’s happening as a part of the development of the smart weapons and virtual presence technology. Robots will definitely be capable of doing more on their own in the not-so-distant future.

Some people talk as though there’s going to be some giant step of autonomous capability in the field. I don’t believe that’s true because when you look back at aviation, changes came in stages. Augmentation systems were followed by auto-pilots, then automatic landings and automatic takeoffs. It was at that point that people realized you can fly an aircraft without a man in the system. We’re seeing the robotic equivalent of those incremental changes today in terms of simple but important things, such as autonomous assistance for the robot operator. Those kinds of technical builds will continue to evolve.

In addition, mission builds for robots also will progress, much the same way they did for aircraft. In World War I, the first air¬planes started out doing reconnaissance, serving as artillery spotters. As time progressed, airplanes were used to perform a range of other missions. Eventually, aircraft were used for their strike capability. It is the evolution of military equipment, and it will come to robots at some point, too. We will build weaponized platforms when our military customers have a requirement based on the needs of our country. But there is an important detail to consider: Our architecture plans always include a man in the loop, and we don’t foresee that changing.

Which brings us to one of the big hot-button questions in the discussion about military robots: Will there be autonomous killing machines? I don’t think so; we’ll never see that life-and-death decision made independent of a human. However, based on the movies, it’s no wonder the American public has a Hollywood-hyped fear of machines taking over the world.

The results of major experiments in the past two years pro¬vide a window on the battlefield of tomorrow — and how unmanned systems will play a prominent role. At the Air Assault Expeditionary Force experiment last October at Fort Benning, Ga., more than 40 types of systems were demonstrat¬ed, including UAVs, UGVs, command-control systems, unmanned ground sensors, hybrid vehicles, lighter-than-air ships and many more. When asked which three systems they would take to war today, soldiers answered: the PackBot UGV and two UAVs. The reason is simple: Soldiers want to get situa¬tional awareness from the ground and the air. UAVs provide a bird’s-eye view, and UGVs provide an up-close-and-personal view. Based on the superior situational awareness they provide while keeping war fighters out of harm’s way, UAVs and UGVs working together will play a vital role on the battlefield of the future.

Joseph W. Dyer is president of iRobot’s Government and Industrial Robots Division. His career in the Navy included positions as the com¬mander of the Naval Air Systems Command, naval aviation’s chief engineer, commander of the Naval Air Warfare Center, Aircraft Division, and F/A-18 program manager. He chairs NASA’s Aerospace Safety Advisory Panel.