Navy Leans on Commercial Technology for Computing Upgrades
When the Navy finally boils down its sprawling information systems into two main networks, they will consist almost entirely of commercially available technology tweaked and retrofitted to meet military security standards.
The dramatic overhaul of the second-largest computing network in the world consists of three main efforts: building an ashore system; its afloat counterpart; and ensuring the two can communicate.
In the IT world, technology progresses at breakneck speeds, often outpacing the Defense Department’s best efforts at modernization. Basing its new network architecture on commercially available technology and rebidding on a cyclical basis for system upgrades is designed to tamp down the cost associated with the service-wide network overhaul.
“This is one of the most commercial-off-the-shelf programs in the Navy’s history,” said Capt. Shawn Hendricks, program manager for the service’s Next Generation Enterprise Network. “It’s all about keeping cost down by leveraging what’s out there and readily available.”
NGEN is the shore-based component of the Navy’s network upgrade strategy. Its shipboard counterpart, the Consolidated Afloat Networks and Enterprise Service, or CANES, is being developed in parallel. Together they will dramatically update and replace the current Navy-Marine Corps Intranet and 11 other networks at sea and ashore. NCMI represents about 80 percent of the Navy’s intranet infrastructure with a user base of about 800,000 personnel, Hendricks said.
While adequate for the service’s current needs, NCMI’s software and hardware components are becoming outdated and upgrades are becoming scarce as commercial technology progresses.
“We’re meeting requirements today, but we’re starting to face problems,” Hendricks said. “I’m running on systems they don’t even make any more. If you can’t get the parts for it, what do you do?”
NGEN will consolidate, upgrade and standardize all the Navy’s shore-based computing and data-storage systems across the United States and at bases and installations abroad. A request for proposals was issued in May. Prior to the July 18 deadline, Hendricks said about 150 responses to the RFP had been received, but wouldn’t elaborate on which companies had submitted proposals. The system is estimated to cost $4.5 billion over four years with full fielding scheduled to be completed sometime around 2016.
“We are still on track for a tentative award date of February next year, but as the circumstances on the ground change, that could change also,” Hendricks said.
NGEN eventually will consist of more than 400,000 computers at 2,500 Navy and Marine Corps installations throughout the world. In a first-of-a-kind move, the services’ command, control, communications and computing systems will be rendered as a secure cloud, hosted in distributed data centers. Many commercial and banking computer systems use the same infrastructure with portals accessing a web of data storage centers. CANES will work in much the same way, but each ship will have its own data center tied into the larger network and linked to shore-based computing facilities.
“That’s just computers,” Hendricks said. “I haven’t even talked about printers and fax machines and telephones. If I service my 800,000 users with 400,000 computers at $1,000 a pop, it’s already at $400 million. This is no small feat. But the idea is not to ask for a whole bunch of new money, because these systems will be upgraded organically with continually bid, commercially available technologies.”
The four-year contract has also been fragmented, allowing the Navy to rebid certain elements in future years. Plans are to schedule a cyclical four-year hardware “refresh” and two-year software refresh that will allow the Navy to keep pace with rapid technological developments. Hendricks said it was a way to keep costs down and an incentive for industry to remain competitive and innovative with their proposals.
“It’s like if you buy a new Ford F-150 and eventually you want to put new tires on it,” he said. “You don’t have to buy those tires from the Ford dealership if you can get better or cheaper ones from somewhere else. In the event we decided to look at another email provider, for instance, we can rebid it in the future and keep everything competitively priced and keep up with technological advancements.”
Competitors can bid on one or both of two segments for NGEN implementation. Within each segment are services the Navy plans to continuously rebid during each technology refresh cycle, Hendricks said. Though he wouldn’t elaborate on companies that have responded with proposals, a Hewlett-Packard/Northrop Grumman team is in the running. HP holds the contract for the current NCMI system.
Northrop Grumman was awarded the contract to develop and implement CANES earlier this year when Lockheed Martin Corp. withdrew a protest. It will soon be installed on an Arleigh Burke-class destroyer and a unit is being developed for installation on an aircraft carrier, said Mike Twyman, vice president and general manager of Defense Systems for Northrop Grumman Information Systems.
Eventually, about 180 ships and submarines and two maritime operations centers will be outfitted with CANES equipment. Here, too, the focus is on commercially available technologies that can be periodically upgraded. CANES is estimated to cost at least $1.5 billion, not including software applications, according to a recent Rand Corp. study.
“CANES is 100 percent commercial-off-the-shelf,” Twyman said. “Using plug-and-play COTS components, for competitive reasons, is the best way to get the best price for the Navy and to allow future technology upgrades. What they really bought is kind of an open-architecture COTS-based design with network-defense technology built in.”
But installing computing equipment on a ship requires more considerations than setting up a data center on land. The environment can stress commercial components while the ship is under way, so technologies like server stacks had to be engineered to withstand heavy vibrations and shock, electromagnetic fields, temperature fluctuations and humidity.
“Those are challenges, but we know how to overcome them and we have,” Twyman said.
Moreover, each component system must be engineered or designed to meet Defense Department security standards. Significant aftermarket security features are built into the software and hardware components of each system to protect classified information.
CANES is nearing milestone-C production and deployment while NGEN is in the proposal stage.
Though CANES is slightly ahead of NGEN, the former may in fact take longer to implement because of ship schedules. Aircraft carriers, for instance, must be upgraded either during construction — as the CVN-78 Gerald R. Ford will be — or during a mid-life refueling overhaul.
“What we’re working on now is aligning ship availability so we can get these things installed,” Twyman said. “That’s proven to be a big logistical task, trying to schedule times when we can get in and do the work while the ships are not out on deployment.”
In development since 2007, NGEN has been criticized as a sprawling program that could result in unforeseen cost overruns. A March 2011 Government Accountability Office study of the program recommended the Navy put the brakes on the project.
“GAO is recommending that DoD limit further investment in NGEN until it conducts an interim review to reconsider the selected acquisition approach,” a summary of the report reads.
Furthermore, GAO found the Navy “does not have a reliable schedule for executing NGEN” and that “the program was approved in the face of known performance shortfalls and risks.”
The Navy “does not have a sufficient basis for knowing that it is pursuing the best approach for acquiring NGEN capabilities and the program’s cost and schedule performance is unlikely to track to estimates,” the GAO report concluded.
Since then, Hendricks and his team of more than 400 uniformed and civilian personnel have shored up some of the vulnerabilities brought up by the report, he said.
“We’ve clarified some areas with regard to software and application hosting and infrastructure and some of the other data requirements,” he said.
Eventually, the Navy’s computing networks will be consolidated into these two main systems, ideally eliminating redundancy and improving security oversight. When both are deployed, the afloat system will seamlessly mesh with the shore-based system. The programs will also standardize and upgrade the physical components of the network, including data centers, computers and telephones. That component will be particularly influential at sea, where some 640 legacy computing systems exist on 300-plus ships.
Some ships have up to 50 networks, each with its own training and support issues, according to the Rand study. Many of those legacy systems are stovepiped and unable to communicate with each other. The Navy has found it difficult to keep up with so many computing networks, allowing for little fiscal oversight and creating vulnerabilities in an age where cyber-attacks are an ever-present threat.
Heavily dependent on robust commercial technology, the new networks are specifically designed to address those issues while minimizing cost.
Commercial technology is also minimizing the risk that the two networks might not mesh as seamlessly as planned. Hendricks said his office is working closely with the CANES program office to ensure the two systems work together.
“We’ve tested interoperability with the shore and across the network,” Twyman said. “CANES design maximizes commonality of components across ship and submarine components. COTS equipment is generally interoperable. Sure, there will be a couple of minor things that could crop up, but there’s nothing huge or any major risks in getting those two networks to talk.”