In case after case, mammoth IT systems that were supposed to improve the management of the Defense Department have overrun their schedules by years, expended billions of dollars more than they were supposed to, and in at least one instance, delivered almost no new capability at all. But the Pentagon says that it’s learned from those mistakes and that among its enterprise resource planning systems, there is good news among the bad.
Among more than a dozen large ERP systems DoD has been working on over the last several years, the Air Force’s Expeditionary Combat Support System has emerged as a favorite target for lawmakers who want to rail against wasteful spending in DoD. At a conference earlier this week, Sen. John McCain (R-Ariz.) decried ECSS, which was supposed to help reinvent the Air Force’s processes for managing logistics and its parts supply chain.
The Air Force canceled the program outright last year after spending more than $1 billion over seven years and getting only about $100 million worth of usable capability.
“The Air Force even managed to pay the contractor over $100 million more between when the Air Force told it to stop work and when the program was officially terminated. And to my knowledge, no one has been held responsible,” McCain said. “What happened with ECSS is a travesty, and in terms of how little benefit was realized compared to how much was spent, it was one of the most egregious examples of mismanagement in recent memory.”
Asked about ECSS at a hearing of the House Armed Services Subcommittee on Intelligence, Emerging Threats and Capabilities Wednesday, Beth McGrath, DoD’s deputy chief management officer, did not attempt to defend the program. But she said the Pentagon had learned from it. “In terms of size and scale, this program was clearly way too big,” she said. “We need to chunk these IT systems into smaller capability sets so that we’re delivering something and then adding to it, instead of trying to deliver the whole thing all at once.”
McGrath said DoD also needs to take more care in preparing its workforce to transition from legacy IT systems and the business processes they operate within to a new generation of ERPs.
She said it also needs to pay closer attention to the issue of data quality.
“For any of these programs, you’re really trying to take data from really old systems and bring them into a new, modern, much more tightly-controlled environment. Also, the importance of infrastructure can’t be overstated in terms of the network, how it runs and how it’s optimized. In every one of these programs, we find too much infrastructure, so it adds to latency and all those kinds of issues,” she said. “We’ve captured all of these lessons learned along with standardization across programs. We weren’t managing or monitoring them in similar way. We’ve made those changes so that we and the program offices can monitor their health across the lifecycle.”
While ECSS is an extreme example of a failed ERP system, it’s not alone in failing to meet the military’s original cost and schedule projections.
Last year, the DoD inspector general examined six systems that it determined would be critical to meeting the Pentagon’s legal deadlines for passing a financial audit. It found all six were years behind schedule, and each of them had exceeded their original lifecycle cost estimates for a collective total of $8 billion in cost overruns so far.
But McGrath said it’s not as though those systems aren’t producing real capability for the department.
“We share both a desire to get this better and a frustration when it doesn’t, but I don’t want to lose sight of some of the capability that’s been delivered,” she said. “We have around 14 of these major business programs. In 2009, we were in a highly-developmental stage. The number of users back then was around 27,000. Today, those same programs have 195,000 users. So we’ve delivered capability. We tend to talk about the programs that are really big and expensive and don’t go so well, but there has been progress made in delivering supply chain capability, financial management capability and also contracting. I don’t want that to get lost.”
McGrath said her office has begun to review the department’s large business systems to nail down the root causes of problems at every stage of the programs’ acquisition lifecycles and look for leading indicators that might show that a program is headed for trouble. She said DoD also will require the sponsors of large business systems to draft a comprehensive business case to justify all of the IT functionality they’re asking for at the very outset of a given program.
Lawmakers are also concerned about the future of an IT program outside the sphere of ERPs: the effort to integrate the electronic health records of DoD and the Department of Veterans Affairs.
The two departments decided to change course late last year. Rather than building one system from the ground up, they’ll each begin from separate starting points and try to incrementally add joint data sharing capabilities.
McGrath said the department’s leadership arrived at that decision after realizing the original concept might prove unaffordable.
“The interagency program office did a bottom-up lifecycle cost estimate that really put the affordability of the approach in question. Secretary (Leon) Panetta and Secretary (Eric) Shinseki asked whether there was a more economical way to still deliver an integrated electronic health record, but in a less-risky way,” she said. “So the goal is to reduce the risk, decrease the cost and maintain the schedule we were on.”
As to the core starting point each department will use, VA will begin with its existing health record system, VistA. DoD wants to eventually jettison its current system, AHLTA, and will make a decision soon about what will replace it.
“We want to ensure we’ve explored all opportunities, so when we’re looking at our core capability, would it be VistA? Would it be something commercial? The health space has made tremendous leaps in terms of modernization over the years, and we want to make sure we’re assessing the capabilities the commercial market brings,” she said. “We issued a request for information in February, and we have all the responses in. We’ll make a determination as to whether we’ll go with a [commercial off-the-shelf] solution or a government solution by the end of March.”
McGrath said DoD and VA are still committed to using a common system architecture and data standards, and she insisted none of the work the two departments have done so far in pursuit of a joint record system will go to waste.
“All the foundational things we’ve been working on since 2011 will all be carried forward,” she said. “We’re not scrapping anything. We’ll continue to use all those foundational pieces.”