QinetiQ North America’s Talon MK-2 ordnance disposal robot
With a chance to build the Navy’s next bomb-disposal robots at their fingertips, unmanned systems manufacturers have an opportunity to tap into one of the only fully funded programs in the autonomous vehicle pipeline. Eager industry officials believe a contract award is imminent.
Whichever company lands a contract to become lead systems integrator for the first of the Navy’s three new unmanned ground vehicles will get a leg up on the competition — losing candidates will have to forage for less lucrative opportunities to build components.
The Advanced Explosive Ordnance Disposal Robot System, or AEODRS, has been alternately touted by the Navy and blasted by industry for its modular, open-architecture requirements, which means one company can’t take the whole cake.
The Defense Department started buying robots in the 1980s by opting for modified commercial-off-the-shelf platforms from different vendors, said Byron Brezina, a technical engineering project manager for the Naval Explosive Ordinance Disposal Technology Division, located in Indian Head, Md. That approach left the military with proprietary systems that are difficult to update and modify.
Upgrading current robots is an “endless cycle” that is further complicated when the military wants to put a sensor or payload from company A onto company B’s robot, he said at the Association for Unmanned Vehicle Systems International’s program review in March. “So you’re seeing those R&D costs, those integration costs being doubled and tripled each time something has to be updated,” he said.
Instead of a single company developing one or all of the bomb-disposal robots, vendors will compete to build one or more of the robot’s components, which will be assembled by the systems integrator of each platform.
The hope is that these robots, which will be the military’s fourth generation of unmanned ground vehicles, will allow the Navy to bring new technology to the field more quickly.
“We’ve learned these lessons … the hard way by having robots in theater and having new threats and requirements come out fast and furiously,” Brezina said. “The open architecture is going to allow us to respond to those quicker, cut down that lead time. It’s also going to help us mitigate technology obsolescence, which is a problem on all DoD tools.”
The first and smallest robot to be developed is a system for dismounted operations, also called increment 1, which is designed for reconnaissance missions. At 35 pounds, it is small enough to carry in a backpack. It has six hours of battery life, a 330-foot range and a manipulator arm capable of lifting five pounds when fully extended, according to Defense Department documents.
The second increment, the tactical operations system, is a medium-sized robot that can be vehicle transported or carried short distances by two people. It should weigh 164 pounds, have a 3,300-foot range and two manipulator arms. At least one of those arms should grasp, grip and pivot, said documents from the Johns Hopkins University Applied Physics Laboratory, systems integrator for the increment-1 prototype.
Increment 3, called the base/infrastructure operations system, should lift up to 300 pounds and would be transported by a large vehicle or trailer.
With the increment-1 prototype in its engineering and development phase, the Navy is seeking approval to release the request for proposals, said a spokesman for Naval Sea Systems Command. Production and deployment of increment 1 could happen as soon as fiscal year 2015.
Increments 2 and 3, which are being developed simultaneously, could be fielded by fiscal year 2018.
The Johns Hopkins lab was recently awarded contracts for systems integration of the next two prototypes, said Michael Zeher, project manager for the lab.
Remotec, a Clinton, Tenn.–based subsidiary of Northrop Grumman, is gunning for the prime systems integrator contract for increment 1, said Phil Bryan, the company’s AEODRS program manager.
Netting that contract would be a boon for the company, which produces the MK-3 remote ordnance neutralization system that is currently in use. Increment 1 will replace the iRobot 310 small unmanned ground vehicle. Increment 2 is slated to take the place of iRobot’s MK-1 PackBot and the QinetiQ North America’s MK-2 Talon.
Early on, the program garnered criticism from industry executives who warned that an open architecture process could lead to higher costs, a distressed supply chain and schedule delays.
The loudest detractors were makers of incumbent UGVs who, unsurprisingly, would rather design and produce an entire system, Bryan said.
That opposition is decreasing, he added. “I think you’re seeing people come around to supporting it. We hope that it’s going to result in more systems being sold and bringing prices down.”
The system design is driven with interoperability in mind. A small, lightweight controller could be used with all three UGVs, and a separate, larger controller would operate the tactial and base/infrastructure operations systems, making it easier to train operators. Navy officials want vendors to create interchangeable parts so that, for instance, a sensor from increment 2 could be put onto increment 3.
The family of three bomb-disposal robots is the first UGV program of record with an autonomy requirement, Brezina said.
For the dismounted-operations system, the robot should autonomously navigate about 10 feet in an unobstructed environment. The tactical operations system and base/infrastructure system needs to “be able to carry multiple tools downrange, and then have the manipulator automatically go to designated tools and change them out for use without having to come back,” he said.
During his speech, Brezina joked that the Johns Hopkins team simply gathered the modules, plugged them in and threw the switch on a working prototype. In reality, developing the robot entailed a long process of making tradeoffs among vendors and repeated revisions of documents.
The increment-1 prototype comprises seven modules such as an electrical power source, vision sensors, a manipulator arm and software that allows autonomy. There is also a communications subsystem linking the UGV with the controller.
Navy bomb-disposal officials and Johns Hopkins scientists defined specifications and detailed the physical, electrical and logical interfaces for each module, Zeher said.
Remotec created the mobility module used in the increment-1 prototype in partnership with MacroUSA Corp., a small McClellan, Calif.-based company specializing in small unmanned systems. The company supplied the chassis and body of the vehicle that other modules are attached to.
The end result was a modified version of one of Macro’s existing UGV designs. Because the systems integrator helps select the team of vendors designing the hardware, MacroUSA will have to compete to design the mobility module without Remotec’s help, Bryan said.
The government provided the electrical-power and autonomous-behavior modules and communications system for the prototype, Bryan said. While all modules will be openly competed, these government-supplied components may create a wild-card scenario where new competitors might have more luck at scoring a contract.
Chatten Associates, a West Conshohocken, Pa.-based company focused on remote vision, developed the visual sensors.
Pittsburgh-based RE2 Inc. — which developed the master, manipulator and end-effector modules — did not respond to repeated requests for comment.
Zeher, Bryan and Chatten Associates President Martha Jane Chatten agreed that satisfying the weight requirement was the biggest challenge in creating the prototype.
The UGV and controller were required to weigh 35 pounds or less and have an 18-inch diameter. Each module also had its own weight limit.
During one stage of development, the Johns Hopkins team found that the cabling between the master and power modules was adding extra weight, Zeher said. The team worked with vendors, eventually deciding to use smaller connectors and lighter-weight cables.
“We lived and died within our weight budget,” Bryan said. “There was certainly a discussion anytime there was any kind of change on the table that would create a weight issue for anybody.”
By the end of the process, the Remotec-MacroUSA team had scaled down the mobility module’s weight from 13 pounds to 11 pounds, Bryan estimated.
For Remotec, another difficulty involved attaching the different modules to the platform it designed. The company worked with Johns Hopkins during the design phase, trading engineering data and models, said Bryan.
“We were basically trying to put the jigsaw puzzle together that gave the optimum configuration of how to put these [components] on here and get the best system performance,” he said. “We had several iterations, and it wasn’t without challenges.”
Throughout the process, the Johns Hopkins team and the vendors used software that is called the systems testbed to simulate whether the modules were properly working together.
“You could have a full system testbed up with no hardware in it at all, and then you could disable one component in the simulation, say the master module or the manipulator … and plug in the real hardware,” Zeher said. “That one hardware component would work with the rest of the system that was simulated.”
Once the vendors delivered their modules, the Johns Hopkins team first would test them individually with the logical, physical and electrical interfaces, and then with the systems testbed, Zeher said. Then it would be integrated with the other hardware.
Using the testbed, Chatten Associates was able to iron out the problems in their visual sensors before bringing it to the applied physics lab to be integrated with other components, Chatten said.
The system testbed software will be freely available to industry. This means that, in the future, any company can develop a component according to the Navy’s minimum requirements and “objective” requirements — a wish list intended to push capability, Zeher said. This will give industry a way to demonstrate to the military that it has an improved technology that works.
Zeher said one of the misconceptions of the program is, because the government defined the interfaces, it wants to own the design.
That simply isn’t what’s happening, he said. “All the government is going to do is focus on what’s at the interfaces and then let industry innovate as much as it can.” Intellectual property will stay behind those interfaces, he said.
“So it really should be healthier for industry in terms of innovation, developing intellectual property and being able to protect it,” he added.
Photo Credit: Navy