|
|
 |
home > May/June 2007 issue > article
|
|
 |
|
| Bottom: IROBOT/Top: Defense department |
|
Robotic evolution
By Brian Robinson
Special to Defense Systems
Robotics technology is evolving to meet the demands of the network-centric battlefield
In the minds of technology experts and military leaders, robotics technology exists in two states at once: current and future.
In its current state, robotics already fills important operational roles. A dozen or so types of unmanned aerial vehicles were used during Operation Iraqi Freedom, and about 1,000 UAVs now fly reconnaissance missions over Iraq and Afghanistan along with missile-carrying Predators.
On the ground, military forces deployed unmanned ground vehicles during and just after the war to search caves and buildings and look for mines. About 5,000 of these UGVs are now on the ground, and they have proved particularly valuable in countering roadside and car bombs.
However, these vehicles are still relatively simple robotic devices. They operate in close cooperation with a human handler, who uses images and other information collected by onboard sensors to evaluate the situation before sending further instructions.
The future state, as described by the experts, looks very different.
On the net-centric battlefield, robots will act autonomously much of the time, using built-in intelligence to guide their actions and seeking instruction from their handlers only when it’s needed.
They’ll also be in regular contact through the network with other robots to coordinate where each needs to be and what it needs to do o carry out a mission plan.
And the more numerous they are, so the theory goes, the more effective they’ll be.
“The underlying thesis of netcentric warfare is that power is proportional to the square of the [networked] nodes,” said Joe Dyer, a retired Navy vice admiral who is now executive vice president and general manager of iRobot Government and Industrial Robots. “So the more nodes you have, particularly those that can move and act as sensors, the better.”
But robotics experts must solve numerous problems before this future can be realized.
The human factor
UAVs are a case in point.
They are the most battle-proven robotic machines, having first seen operational success in the Vietnam War and with Israeli forces in the 1970s, yet UAVs are still flown remotely by human operators. Even the relatively mature high-altitude UAVs are remotely controlled.
UGVs have only been around for the past 10 years, although, with the success they’ve had in Iraq against improvised explosive devices and in searching buildings, they are expected to play a large role in the future.
There’s tremendous excitement about the use of robots as an extension of the infantry, Dyer said. Clearing a building by conventional means takes cautious, SWAT-style tactics, he said, whereas you can “capriciously send robots in to see what confronts you.”
UGVs also help to speed tactical operations. “If the soldiers see that there’s nothing ahead of them, they can move quickly to re-establish their position instead of having to be cautious at every step,” he said.
But the human factor is a limitation. The military needs to get away from having one operator for every robot, said Bill Thomasmeyer, president of the National Center for Defense Robotics. That means integrating more intelligence into the machines so they can do more on their own, allowing one operator to control a number of robots at the same time.
Ideally, for example, an operator might guide a robot to an IED but then leave it to disable the IED and get back to base on its own, while the operator guides another robot to disable another IED or perhaps search a building.
“The computing power and algorithms don’t exist yet to provide full autonomy,” Thomasmeyer said. “But the technology does exist for some degree of semiautonomy.”
Managing the data
For example, Insitu, which develops high-endurance UAVs, recently chose commercial messaging middleware from Real-Time Innovations for its next generation of ScanEagle and other UAVs.
RTI’s distributed-data software uses a publish-and-subscribe method to move data between ground control stations and the UAV. Data is only provided to those systems that need it, not every system on the network that may or may not use it.
The result is a much more efficient use of data. It provides a better-orchestrated dataflow that allows it, among other things, to switch easily among multiple ground control stations and to connect to a UAV reliably, even via unreliable links.
“In net-centric communications, you are always limited by the realm of physics and what you can and can’t do over a radio link,” said Gordon Hunt, principal radio applications engineer at RTI. “This provides for dynamic discovery and essentially decouples the data from the hardware.”
The Army and Navy have adopted data distribution systems for use in their future net-centric programs.
Data distribution also provides for much more efficient use of
bandwidth, which is probably the overriding concern in a net-centric environment.
The bandwidth bottleneck
Bandwidth will be in high demand in the Army’s Future Combat Systems environment. FCS is the Army’s vision of the net-centric battlefield. It entails soldiers working seamlessly together with manned and unmanned vehicles as a single, cohesive system of systems in which the capabilities of the whole are greater than the sum of the individual parts.
More than a dozen robotic systems — ranging from small ones that can be carried in a soldier’s backpack to large gun platforms — are included in FCS.
“But the communications needed for this will require a lot more bandwidth than we currently use, and there is only so much available in any one area of operations,” said Lt. Col. Steve Noe, product manager of FCS (Brigade Combat Team) for unmanned ground vehicles. “We have to be able to deconflict that, and we are still trying to analyze how many vehicles at any one time can operate in a specified space.”
A lot of work on this problem is being done through both FCS and the Joint Tactical Radio System program, he said. JTRS is an ambitious program to replace aging battlefield radios with a family of software-programmable tactical radios that can operate across a much broader set of frequencies.
A key consideration in robotic communications is the need to get information produced by their sensors into the handheld/radio controller of the warfighter on the ground. The farther away a robot is, the greater the demands on the communications system.
But designers must weigh those demands against the size, weight and power requirements of the overall robotic system, which could require trade-offs among the various components, Noe said.
Emerging technologies need to be hardened and made more reliable, then transitioned into operational capabilities, Thomasmeyer said.
“You really need to get these technologies into the combat schools and into the hands of the developers there,” he said. “People are starting to become intrigued by the possibility of these robotic technologies, but they don’t yet fully understand how they can use them.”
Noe said the prototypes his organization is developing should begin their initial qualification process within two years.
|
|
|
|
![1105 Media [valor]](/images/ds1_pntmlogo.gif "1105 Media [valor]"){kind=small}
