BOTS
Written by Adam Baddeley
USSOCOM is one exemplar of this trend, launching its new Combat Autonomous Mobility System (CAMS) joint capability technology demonstration in fiscal year 2008. CAMS is a modular robotic vehicle solution whose tasks encompass a range of battlefield roles including re-supply, joint fires support, extended reconnaissance and perimeter security in support of multiple SOF mission profiles.
“Autonomous mobility provides a transformational way for ground forces to operate in the modern day battlefield,” said Scott Forman, CAMS project manager, USSOCOM Advanced Technology Branch.
General Dynamics Robotics Systems was awarded a $10 million contract for a three-year phase of work in December, which will see modules being integrated onto a number of different governmentfurnished vehicles for testing by special operations forces.
Newcomers to the UGV user community will inevitably play catch-up. “The EOD guys get it; they get it well. The concepts of operations have been developed, and the tactics, techniques and procedures are in place,” explained Bob Quinn, vice president of Talon operations for QinetiQ North America’s Technology Solutions Group. “On the combat arms they are back to where IEDs were prior to the beginning of the war in Iraq. The bottleneck isn’t with the technology; the bottleneck is with the willingness to change the way operations are conducted today and willingness for combat arms to use small robots in their everyday operations. The necessary technology that has been demonstrated—additional levels of autonomy have been developed but not yet implemented.”
Some things of course remain the same, namely UGV’s inherent expendability as Alan Bignall, president and CEO, ReconRobotics Inc., explained, “There is a distinct value proposition in sticking a set of electronic eyes, rather than your head, around a corner.”
EOD AND BEYOND
Changes are also coming in EOD too. The Department of Defense is coalescing its EOD requirement into the next-generation EOD robot, led by Naval Explosive Ordnance Disposal Technology Division, Indian Head, Md., with a solicitation scheduled for 2010. This will see units equipped with three classes of robots: small, medium and large.
Quinn commented, “It is abundantly plain to users that there is no one-size-fitsall and that you really need different-sized robots for different missions.”
The weight of today’s EOD UGVs means they are effectively limited to being carried by another vehicle. However, MRAPs, the vehicle of choice for EOD, are effectively road bound, limiting where current EOD UGVs can be easily deployed into position. With the “small” option, EOD platforms can be easily man-packed by troops.
The move is not so much a change in tack, rather a continuation of the evolution of UGVs. Quinn commented, “If you look at the technology road map, it began with the DARPA program in the late 1990s to do two things; to taking big, heavy, slow robots and making them smaller, faster and manportable but also a capability for CONOPS to be able to work with the human operator.”
The second phase of UGV development is seeing the proliferation of smaller robots in the hands of combat arms. Quinn said, “That is where small robots like QinetiQ’s Dragon Runner come into play for combat arms applications, where people are actually using the robots not as a counter-IED device but rather as a surveillance device.”
The Dragon Runner began life as a U.S. Marine-funded project with a Carnegie Mellon University firm, buying 35 in total but just one or two at a time. QinetiQ bought the company in 2007. Quinn said, “We liked it because it’s easy to operate, rugged as a son-of-a-gun and fast. It met all the criteria of our Talon, but it was much smaller.”
He distinguishes between Dragon Runner and the company’s Talon UGV, used extensively by combat engineers for route clearance, with more than 2,500 in Afghanistan and Iraq. When using only a smaller robot in counter-IED missions, “You lose the counter-IED mission from the standpoint of using manipulators to remove debris from around the IED,” Quinn said. “Talons pull, twist and rip off the initiators to bring them back as forensic evidence. Talons have also dragged downed insurgents with suicide belts right out of a building as well as pulling donkeys, cows and animals who have been stuffed with explosives off the road. That’s very hard for small robots; the physics get in the way. However, soldiers storming a building don’t need a bigger robot.”
The Dragon Runner has already been put in the hands of U.S. and coalition partners in Iraq and Afghanistan with a focus on the latter. Quinn said, “The expectation is that the volumes will be quite a lot larger than for the talon systems.”
Field changeable options are vital irrespective of the size of platform. The Dragon Runner uses a Picatinny rail as the interface to which sensors can be physically added, using a “plug and play” architecture to electronically integrate their inputs.
Quinn cites the experience with the Talon to illustrate how important this is. “Our former military employees will tell you, no one mission is identical to the other, so having field transformable options and accessories, which can be applied without hand tools based on the situation in hand, are very useful to soldiers.”
Other advantages can be derived by improvements beyond the platform such as forward fielding of robot repair facilities. However, Quinn believes the biggest technological change is the throwing away of the operator’s control unit in a suitcase in favor of a more durable, wearable solution, closer to a laptop weight.
POCKET ROBOTS
While dedicated EOD robots and their spin-offs had a well-established presence in the market dating back to the 1970s, their mini sub-kilogram cousins are much more recent additions to the inventory and until recently were only pursued in the research community. The goal of these miniature robots is to give combat infantrymen, rather than specialists, extremely rugged, ultralightweight systems that can be thrown into an environment and then monitored and directed using intuitive controls.
This emerging capability is reflected in the genesis of the Recon Scout platform developed by ReconRobotics Inc. A spinout from the University of Minnesota, its underlying research expertise in robotics led to the receipt of funding primarily from DARPA but also the National Science Foundation for man-portable robots under a DARPA program in which 30 working prototypes were distributed to military units in Iraq and Afghanistan.
Bignall explained, “The technology was created and patented at the University; we then spun the technology out to the new company, and the three inventors from the university became full-time employees.”
The first platform, the Recon Scout reconnaissance robot, reached users in mid- 2007, with approximately 300 in use around the world, 25 percent being international customers. In the United States, they are being used in the full spectrum of operations from police SWAT—the largest user—to counter-terrorist and military battlefield operations.
Bignall said, “We’ve focused on police use first. It enabled us to pursue our strategy of getting a hundred different units using one robot as opposed to chasing big military contracts.” That is now changing. Bignall continued, “The military has become aware of the Recon Scout, and they are acquiring them on a unit-level basis as well as seeking variants for additional mission solutions. We believe that will then attract the larger number orders.” In terms of military user he added, “The focus tends to have been on special forces in the early stages, for obvious reasons.”
The UGVs themselves are a “dog bone” or “dumbbell” design. A small “tail” is added to maintain the alignment, ensuring the camera is always looking forward. The shell is titanium. The wheels are a custom polymer and function as the shock absorbers taking 1200–1500 gs when the robot is repeatedly dropped from over 30 feet onto concrete. Inside the shell is a chassis to which components are bolted. The systems weigh just one pound—all in.
Bignall explained the basic principles of operation. “Its whole concept is as a portable, throwable and mobile quick-use reconnaissance robot that can transmit real-time video indoors up to 100 feet and outdoors up to 300 feet. You can easily throw it 50–100 feet, bouncing down the street. Right now it is only equipped with video sensors operating in black and white because it operates in very low light. The minute you go to color, you cannot go to that level of low light capability in our miniature configuration. Our future is very clearly going to add other sensors to that capability, such as chemical and biological.”
As users get a hold of the UGV, features have been added, based on their feedback. The latest version, the Recon IR, for example, has had a light bar added to operate in absolute darkness and has been well received by the community. Bignall said, “When the camera senses it is too dark, the IR emitters will turn on. It’s all automatic. We have a philosophy of no buttons or dials.”
This minimalist approach is reflected in the control unit, which has one joy stick and no buttons or switches. Bignall explained the results, “There is no training, you pull the pin on the robot, which is the on-off switch and throw it. There is a toggle on-off switch on the control unit, and then all there is the joystick.”
Ultimately, this robot is all about giving the soldier mission-critical information in real time. This information protects the lives of the soldiers and enables them to complete a variety of missions quickly and effectively.
TAKING A LOAD OFF
The combat loads of combat soldier has risen over the last 10 years, and all involved are seeking to identify ways of lightening the load that don’t interfere with the mission but actually complement and enhance effectiveness so that when soldiers and Marines get to the culminating point, they can still fight.
The best known of these programs is the Multifunction Utility/Logistics Equipment Vehicle (MULE), for which Lockheed Martin is the prime. MULE is a part of the Future Combat Systems program. Don Nimblett, business development manager for unmanned ground systems said, “The MULE is designed to support infantry in dismounted operations but specifically, heavy or mechanized infantry. It is designed to carry two infantry squads worthy of gear—about 20 personnel. It is very well suited to that. We have one vehicle, a pre-prototype that we have been experimenting with to better understand the complexities of its mobility.” Under the program, Lockheed Martin builds the MULE mobility chassis and integrates the General Dynamics Robotic Systems autonomy and navigation system.
Independently, Lockheed Martin began developing the Squad Mission Support System (SMSS) for special forces and light infantry. Nimblett explained, “We decided that they could use a vehicle like the MULE but that it would have to be smaller. We basically conceived the idea of a one-squad-sized robotic vehicle versus the two squads with MULE. We started in 2006 and have evolved the system to the point that we have just finished providing two SMSS vehicles to the U.S. Army at Fort Benning for the Army Expeditionary Warrior Experiment.”
The SMSS vehicles were given to a light infantry platoon who operated the vehicles for a month through 10 scenarios. At the end of that an evaluation was made as to the merits of the technology against the Army’s needs. The scenarios covered the full spectrum of operations. Seven were night missions. They included an attack of a small city—the McKenna MOUT (military operations in urban terrain) site—to find hostages and clear the area of the insurgents, ambushes in forested areas, and reconnaissance missions.
Prior to Fort Benning, the SMSS began its testing at Fort Knox in Kentucky and has been operated in mountain foothills in Colorado.
The basis for the SMSS is a Land Tamer all-terrain vehicle to which Lockheed Martin added military ruggedization and installed its autonomous capability.
There are two autonomy paths in the SMSS’ programming. Nimblett explained, “The first is to use one of its sensors to find the person it is supposed to follow, create an image of that person and then follow them. There is weakness with that, which we recognize as sensor technology [that] doesn’t let you specify who that individual is. All it sees is a human figure. If you were, for example, leading it and I crossed behind you, it might start following me. Until we can develop the sensors and algorithms that allow the system to identify a specific person, we have another autonomy system in the controller that the individual carries. It sends out a signal to the vehicle to tell it where it has been and leave what we call an electronic breadcrumb, detailing that route, which it then follows.”
The SMSS can’t always follow the soldier’s route everywhere. When this happens, it falls back on additional options. There are four ways to operate the vehicle; first and foremost soldiers can get in and drive it, using the driver’s station. The second way is traditional remote control. The third way is tele-operation, using your control unit display. Nimblett explained, “You tap into the sensor on the vehicles and you look at those sensors and you drive the unit through that means. Unlike the RC option, you don’t have to be able to see the SMSS to operate it.”
“The fourth way is the autonomous capability,” explained Nimblett. “You can tell it to go to a point and when you release it, it will find its own way to that point. A second way is that you literally pick out the route for the vehicle on the display, and tell it to follow that route exactly. Then we get to the ‘follow-me capabilities.’ SMSS has the ability to retrace its route. Right now, if it loses full communications, it just stops for safety purposes. However, we also give it the capability— should we want to do that—where if it loses communication, it essentially turns back and starts back on its route until it re-establishes communications with the operator.”
In and among all that capability, there are sensors on board that allow the vehicle to see and avoid obstacles. Nimblett explained, “It will avoid buildings, and if it comes to a wall that it can identify and it knows what it cannot climb, it will stop, search through it on its on-board maps system and see if it can find another route to get around this and then proceed on that route. If it can’t figure it out, it then notifies the operator that it needs help. In the past, we’ve watched other robotics systems try to go through a series of obstacles and get confused and then waste valuable time trying to find a way out. We figure the best way of doing that is recognizing it’s going to happen and program the SMSS to alert the operators that it needs help.”
Light and special forces rely on stealth. Nimblett said that feedback from soldiers showed the current turbo diesel solution is noisy and Lockheed Martin is addressing that. “The short-term fix will be to dampen down the engine noise and the hydrostatic propulsion system with insulation and special mufflers, which is the noisiest part of the vehicle, but ultimately we are looking at a hybrid electric system.”
Theft from any cargo UGV, either by thieves or children, particularly in urban areas, is a recognized issue. “We are working with the customer on this. We are looking at designs where the cargo area could be covered to reduce pilferage; somebody reaching in and grabbing a box of ammunition or taking someone’s rucksack, but at the same time allowing quick access by the soldiers. This vehicle in our opinion is a squad support vehicle; it is going to be very close to the soldiers in most cases, acting as a deterrent.”
A second build of vehicles is now under way with a fleet of seven expected by late spring. In addition to cargo haulers, a platoon level reconnaissance variant and civil fire fighting vehicle have been designed. These vehicles will be capable of being a slung load under an MH-60, and two vehicles will go internally into a CH/MH-47. An armed reconnaissance capability was considered but has now been dropped based on feedback. “In talking with the customer they are not quite ready for an armed robot to be in amongst the squad.”
The SMSS’ autonomy package is platform agnostic, although Lockheed Martin plans to continue with the Land Tamer for the foreseeable future. The same system has also been used to control a four-strong cargo truck convoy for over 100 miles. Nimblett cites the key UGV sensors as LADAR, radar, TV camera and FLIRS. The real differentiation he explained is in the sensors. “The heart of it is who can best develop the most effective and efficient algorithms to interpret the data that is provided by the sensors.”
Lockheed Martin is now co-operating with the Army and Marines. The Army plans to take four vehicles with a light infantry and potentially deploy to Afghanistan in April– May 2010. In June, the Marines will take a further two platforms for a similar, six-month limited objective experiment. One vehicle will operate as a squad support vehicle, the second as a resupply vehicle, traveling up to 30 kilometers to remotely re-supply a unit.
Nimblett said, “If it is all successful, then I think the Army and the Marine Corps will have sufficient data to develop the requirements documents that will begin the acquisition process. I would anticipate seeing requirement documents developed in late 2010, early 2011.” ♦