Commentary

Unmanned but Not Untethered

Robots on the Future Battlefield

By John Matsumura and Randall Steeb

John Matsumura is a senior engineer at RAND. Randall Steeb is a senior computer scientist at RAND.

When a Predator drone fired a Hellfire missile at a car and killed six suspected members of al Qaeda in Yemen in 2002, the direct hit dramatically highlighted the potential for using robots during war. But it also raised the specter of robots replacing human beings and, for some, conjured up images of future armies of Star Wars “droids” running amok on the battlefield.

At this point, robots fighting without humans in the loop is science fiction; but using robots on the battlefield is a reality, one that received a huge jump start from a mandate in the fiscal year 2001 Senate defense authorization bill that “by 2015, one third of the operational ground combat vehicles of the armed forces will be unmanned.” Growing numbers of robots will clearly move across and above future battlefields, but the robots will augment — not replace — soldiers, performing missions that can best be described as “the dirty, the dull, and the dangerous.”

Future battlefields, potentially contaminated by harmful or deadly toxins, could become much less suitable for humans. Taking advantage of the inherent robustness of robots in such “dirty” environments could not only compensate for shortfalls but also provide new capabilities. For example, robots could be used to mark contaminated areas or to neutralize dirty areas (by means of biological cleanup).

Robotic technology that largely already exists can free up manpower by accomplishing many of the “dull” tasks of military missions — tasks such as hauling supplies — that now employ soldiers simply because they may be the only readily available resource. And emerging technologies, such as enhanced voice recognition software and adaptive route planning methods, can expand the range of the “dull” missions into areas where it is just more practical to use robots, such as in providing battlefield surveillance.

Robots also offer the promise of helping to keep soldiers more out of harm’s way. But such “dangerous” missions — where weapons are fired during the “fog of war” and when friendly forces and noncombatants are in the mix — require an autonomy that robots do not yet possess. Despite progress in artificial intelligence technology, we still have a very long way to go before robots can replicate the human thought process in such complex situations. The best strategy for the foreseeable future is to integrate soldiers and robots so that the whole is greater than the sum of the parts. Although the Predator was uninhabited, it was very much “manned” in the sense that a soldier controlled its every move, seeing what it saw and, ultimately, pulling the trigger.

Concepts for how robots can be used in dirty, dull, and dangerous situations are rapidly developing, but it is unclear how practical and feasible such concepts are. Critical pieces of concept exploration, technology development, and force analysis do exist. But given the mandate to have a substantial number of unmanned systems in the force by 2015, the research, analytic, and development communities are sure to be tested in the coming years.

What makes sense now is to organize the core challenges — technical, tactical, and strategic — into an integrated framework that can be quantitatively explored and assessed. Although many of the analytic models and simulations needed for this task are already in place, many more will have to be developed. The key variable for success over the next few years will be how well the research, analytic, and development communities can work together to shape what will surely be a revolution in military affairs.

PHOTOS: DIANE BALDWIN