Urban Challenge laid the foundation for robotic convoys and other wheeled or tracked logistics solutions in the future. That includes scenarios in which robots would move through a busy marketplace with a squad on patrol, carrying equipment and helping with surveillance. That would require an advanced perception and planning system capable of dealing with sudden and unexpected movement all around it.

For most combat operations, the ability to walk remains the preeminent goal of robotics. Ongoing combat operations in Southwest Asia add to the urgency.

“The LS3 program will run three to four years, after which we look to transition it to the Marines or Army. Realistically, 2015 is probably about the right time frame for it to begin fielding,” Mandelbaum predicted. “Legs for the future of robotics sounds a bit far-fetched to some – until you see videos of Big Dog [viewable on www.youtube.com] – but they also are the major solution developed by nature over millions of years of evolution.”

As more and more robots enter into military service, another factor will be pushing them toward greater autonomy: Limited battlefield bandwidth.

“Communications is a major issue. You’re talking about a series of platforms now being operated 99 percent of the time by the warfighter using the robot, all teleoperated or remote control, with very little if any autonomy,” Overholt said. “The problem on the modern battlefield is bandwidth, with several systems all connected and trying to use limited bandwidth to pass information back and forth. That bandwidth has to be robust or your ops tempo suffers. We are starting very slowly to integrate autonomous capability into these robots to reduce bandwidth requirements.”

Another issue, growing in importance as U.S. military operations in Iraq draw down while significantly increasing in Afghanistan, is meeting the often radically different needs of different combat environments.

“Robotic convoying was in response to moving goods and services around Iraq, where the convoys were targets and we wanted to help the soldiers by enabling them to take their eyes off the road,” Overholt said. “Convoying will still be important in Afghanistan, but there are other issues we must address. There are no roads and we have to hump these robots up some fairly steep hills and mountains, so making them lightweight is critical.

“As we continue the fight there, new information will be coming in and we will have to respond in kind. And if we go back into a jungle environment, we’ll have a whole new set of requirements. We have to respond to the environment and terrain in which the warfighter must work.”

There also are potential applications some might consider to be straight out of science fiction, such as sending robots ahead of warfighters to build base camps.

“That’s on the list. I have an idea for a standardized box of robots we configure on the plane and drop into a battle area and they set up the base for the warfighters. That is exactly what NASA is doing for deep space. When we go to Mars, robots will go out first, set up the habitat and a refinery for fuel when the astronauts show up a couple of years later,” Braden said.

“So that is not out of the question. But we also try to stay grounded in what the soldier and Marine needs now. That is a major tenet of what the Joint Center does now – what do they need right now. But we also spend some time looking into the future, trying to envision what we could be doing with robots.”

To Mandelbaum, the real goal of combat robotics is to change the very nature of war itself.

“The first applications of robotics will be things to help ground soldiers and the rest of the battle will remain the same,” he concluded. “But my view of the future of robotics – and I hope to get the right programs into DARPA – is to change the way we fundamentally think about battle. I can’t be more specific about particular programs, but you can think in terms of communications, remote surveillance, the ability to do things at a distance.”