The vest, which weighs about 10 pounds, carries a battery, the handheld controller and a tablet that flips down from the soldier's chest so he can see what the robot sees. That means the soldier doesn't have to see the live target himself.
The robots also can be operated, via satellite radio communications, from hundreds of miles away.
As for security, executives from the robotics companies said there is concern about protecting the robots from being stolen or hacked. Several companies said that to ensure a hacker can't break into the robot's computer system and turn the its weapons on U.S. troops, the military relies heavily on encrypted radio systems.
The idea is for the robots to move along with the troops, by either being air dropped or traveling with the soldiers as they patrol on foot.
The weaponized robots could offer the soldiers heavy fire power backup, which would be particularly important when larger guns or tanks aren't available.
A robot could move out into an open area, using its sensors to detect enemy combatants, while human soldiers stay safely behind. A robot also could move into position a greater distance from its squad, so the enemy will be focused on defending itself from the robot while U.S. troops move in on them.
"It's actually a good thing," said Staff Sgt. Douglas Briggs, Maneuver Battle Lab NCO, stationed at Fort Benning. "It keeps soldiers out of harm's way."
Briggs, who has worked with robotic-armed machines equipped with machine guns in Iraq, said a big part of bringing robots to support active duty soldiers is trust. The soldiers need to trust that their robot will not only work when needed, but will not harm them.
"It comes back to old ways and incorporating new stuff," Briggs said. "We need to see if it's going to do what they say it will do. It's like when we started using GPS instead of a compass. I trusted my compass. I had to get used to GPS."
While robots may eventually become trusted members of a squad, the military is far from willing to give a robot the autonomy to fire on its own.
Tollie Strode Jr., a senior project officer with the Maneuver Battle Lab at Fort Benning, said, at this time, there always will be a human in the loop when a decision is made to have a robot fire a lethal weapon.
"The robot may acquire an enemy target, but it will still always ask a human for permission to fire," Strode said. "I think the ability for a robot to acquire and assess a target and ID it as a threat and fire is probably five or 10 years out. However, even if that capability exists... we'll have a human in the process of deciding what to do."
Sign up for Computerworld eNewsletters.