It is no secret that robots are rapidly asserting themselves in today's military applications. For the duration of the Iraq War, robots have roamed the streets of Baghdad, Fallujah, and other Iraqi cities. And, last summer, we saw the introduction of the first-ever armed robots being deployed into battle — the "special weapons observation remote reconnaissance direct action system" (SWORDS) robots come equipped with M249 machine guns and were designed not just for reconnaissance, but also warfare.

These robots, however, have yet to fire their weapons because of concerns over friendly fire and civilian casualties. The Army, interested in the ways that robots could aid in fighting, has yet to become comfortable with the security of using robots in more than a support and reconnaissance capacity.

Part of the reason for this concern has to do with the way that robots are being controlled on the battlefield today. Historically, robots have been controlled from afar, by a pelicancase- based controller. In this scenario, robot operators are stationed remotely and control the robot using cameras, audio feedback from a fire team, and other cues. While this method has the advantage of keeping the human operator away from the dangers of the battlefield, it severely limits the operator's ability to understand what is happening on the ground and make rapid tactical decisions. It is this disconnect, at least in part, that is responsible for much of the concern the Army has regarding robots being used for shooting or other quick-action tasks. A better option for robotic control — and one that has gained traction in recent military tests — is for the operator control unit to be worn by a member of the fire team. This approach, though not without limitations, provides the operator with a more up - front perspective and the ability to make safer and more accurate decisions than the remotely-positioned operator.

Test Case for Man-Wearable Control Units

Last summer, the Army Robotic Systems Joint Project Office (RSJPO) hosted an event at Fort Benning that was instructive regarding the future of military robotics. Formed as a joint venture between the Army and Marine Corps, the RSJPO showcased a pair of armed robots and explored their usefulness as assets to an assault fire team.

Figure 1. The Army’s MAARS robot, developed by Foster Miller.
The first robot was the Army's MAARS robot (Modular Advanced Armed Robotic System), developed by Foster Miller (see Figure 1). The MAARS robot leveraged the traditional approach to operator control, stationing a remote controller away from the fire team. While the robot was effective in achieving many of its objectives, the remotely stationed pelican- case controller revealed a couple of limitations. First, one member of the fire team had to be stationed remotely. This meant that this fire team member could not be on the ground to coordinate mission specifics and work with his team. It also meant that the controller was stranded in another location — theoretically, if not practically, alone and in danger as the rest of his fire team was completing a mission. The second issue was the communication lag that existed between the fire team members close to the robot and the operator stationed remotely. Like an elementary school game of telephone in which students struggle to maintain the meaning of a message as it is passed from one student to the next, the operator had to interpret the suggestions and comments from other fire team members on the ground. These two problems signaled not a problem with the MAARS robot, but rather, the manner in which it was being controlled.

Figure 2. The Marine Corps’s Gladiator robot, developed by Carnegie Mellon.
The second robot being evaluated was the Marine Corps's Gladiator robot (developed by Carnegie Mellon), shown in Figure 2. This robot leveraged a man-wearable control unit in which a member of the fire team was able to directly control the robot via a system worn on his vest. Not only did this improve the speed and accuracy with which the robot could be controlled, but it also improved the cohesiveness and effectiveness of the fire team. The controller, no longer stationed remotely, could perform other duties as part of the fire team when not busy controlling the Gladiator robot.

The Role of a Robotic Squad Member

The advantages of having a man-wearable control unit only apply insofar as the robotic squad member provides a certain level of service, safety, and advantage. As is alluded to above, a man-wearable control unit offers an embedded fire team a number of advantages over a remotely positioned pelican-case controller. Soldiers are free to remain as part of a cohesive fire team, the cadence of operations is not impeded by communications back to a remote operator, firepower is im proved because of the local operator's improved perspective, and the setup and movement of the control unit is considerably easier when it is appended to a soldier in the field. However, these advantages come with a few detractions; namely, increased risk to the operator in the field, operator loading, and the increased burden of protecting the operator while the robot is being controlled. Thus, while the man-wearable control unit is proving to be a superior choice than a pelican-casebased control unit, there are several key characteristics that the controller must adhere to.

For the robot to be beneficial to a fire team, it must provide the team with an advantage without handicapping the fire team or compromising the mission. There are potential conflicting levels of importance on freeing up the soldiers and enabling the robot to take an increased role in surveillance and firepower. In one way, the man-wearable control unit frees up the operator when compared to a pelican-case-based controller. No longer stranded remotely, the operator is a more effective participant; but, controlling from the field means that, when controlling, the operator is somewhat distracted. This leads to handicapping: With all of the value of a manwearable control unit, care must be taken to ensure that the benefits (improved effectiveness of both the robot and operator) are not outweighed by negative consequences (increased danger due to operator distraction).

Beyond these elements, the concerns pertaining to the robotic squad member are the same as they would be for a remotely controlled unit. Servicing, transport, and reliability are likely to be similar in both man-wearable and remote-location control scenarios. While the argument could be made that man-wearable control could provide efficiencies in these areas because of the ability to directly communicate with the robot, these advantages are likely to be few. The primary concern regarding robotic control in theater, then, remains the safety and effectiveness of the fire team.

Requirements for a Man-Wearable Control Unit

In order to minimize danger to the fire team and improve effectiveness, the robotic control unit should be easy and natural to use, thus minimally burdening the operator — both physically and mentally. Weight is a key concern. Pelicancase controllers are impractical for manwearable scenarios because of their excessive weight; such a controller would literally weigh down the fire team member to an unacceptable extent. The ideal control unit, then, should weigh no more than a couple of pounds. At this weight, the soldier can easily wear the system on his or her vest without negatively impacting mobility.

Figure 3. (Left) iRobot’s Small Unmanned Ground Vehicle (SUGV) leverages Quantum3D’s Thermite Tactical Visual Computer for command and control (right).
Reliability is another key concern. Identifying a lightweight processor that is sufficiently ruggedized can be a difficult proposition. Not only should the processor be enclosed in an indestructible casing (to prevent it from being damaged as the soldier crawls or drops to the ground), but it also needs to be sufficiently protected from the elements. This means that the unit should be rated for extreme temperatures, be sand- and dirt-proof, and be capable of running in wet conditions. This balance of ruggedness and weight requires a different approach. Even traditional "tough" laptop computing solutions are insufficient. Robotic control requires an embedded system that is specifically designed for deployed military applications, such as the Thermite Tactical Visual Computer shown in Figure 3. Such a computer is necessary for reliability. All of the graphics capability in the world is worthless if the processor is incapable of performing in the field.

No less important for the control unit is a long battery life. The objective is to ensure continued operation for prolonged periods of time. And, finally, the system must deliver clear graphics and video capabilities, without which the system's convenience would be made irrelevant.

At the beginning of the article, we noted that the only armed robots in Iraq — the SWORDS — have yet to fire their weapons in battle. This has to do with the fact that it is difficult to make fire/no fire decisions from a remote location. The rounds from a submachine gun will travel through multiple bodies or walls, making the decision to fire a potentially devastating incident. The decision to fire is something that soldiers are trained for, and something that a remote operator, with limited information, is under-qualified to determine.

The soldier on the ground, however, has been trained to make these decisions and has a more accurate perspective of the robot's position. The operator with the man-wearable control unit has not only the training, but also the positioning to make these types of critical determinations.

It is these types of situations that ultimately point to the power of a man-wearable system. While caution needs to be exercised and procedures put in place to protect the field-embedded operator, the value derived from having an embedded operator far outweighs the risks. As with the fire/no fire decision, the embedded operator is able to maximize the value of what the robot was initially intended to provide — battlefield assistance that improves soldier safety and increases the likelihood of mission success.

This article was written by Rick Davis, Principal Application Engineer for Quantum3D, San Jose, CA. For more information, click here .

Defense Tech Briefs Magazine

This article first appeared in the February, 2009 issue of Defense Tech Briefs Magazine.

Read more articles from this issue here.

Read more articles from the archives here.