Retired IT specialist shares inside story of botched National Parks Moose project
The table setter when I arrived was a humorous/informative look by Ekahau's Jussi Kiviniemi at designing Wi-Fi networks for high capacity. The presenter compared such network installation and design to that of setting up a bar, but also made pointed observations about the conference center’s own imperfect Wi-Fi installation history.
The next presentation (“The Moose Project: What Went Wrong An ICT Case Study from the National Park Service”) was as fiery a talk at a tech conference as I’ve ever heard. Recently retired National Park Service IT specialist Michael Thornton emphasized that he didn’t want to “bash anybody or point fingers” over what he described as a systemic problem with architectural, engineering and construction (AEC) projects, but at the same time he is urging fellow members of the information and communications technology field (ICT) to rise up and convince organizations that ICT pros need to be included in project plans from the start – or else risk botching those projects and wasting millions of dollars.
MORE: And BICSI's 2016 Cabling Installer of the Year is...
Thornton, a Registered Communications Distribution Designer (RCDD) who has designed, installed or managed more than 100 ICT projects in Yosemite, Denali and other national parks over 20 years, began:
“How many of you have been frustrated by communications over ICT projects Frustrated with the rest of the project team, frustrated with the [architecture & engineering (A&E)] team How many projects have you had where you came away shaking your head and thinking man if they had only brought us in sooner If only they had brought us in sooner we could have given them a much better solution for this project in their project budget.”
And so on. And so on.
Such problems have “plagued us in ICT since the beginning of the modern age of computers,” the 20-year ICT veteran said. A&E workflows have been ingrained within the cultures at some organizations since before ICT even emerged, he said. It’s up to the ICT industry now to help these organizations help themselves and get out of their rut, said Thornton, describing his presentation as “an objective critique.”
To illustrate his point Thornton took attendees behind the scenes at one federal civilian agency, the 20,000-employee National Park Service, to show how it manages A&E projects. More specifically, he took us inside the Moose (Wyoming) headquarters of Grand Teton National Park, which like the other more than 400 national parks and sites relies heavily on information technology in offices and in the field – “and that means they’re all heavily reliant on ICT, doesn’t it”
As he pointed out, NPS real estate includes some 24,000 buildings, at least a third of which by his estimation are equipped with computer/network infrastructure of some sort, including some increasingly multimedia-heavy visitors’ buildings. The Park Service is connected nationally by a WAN farmed out to Verizon Business Network Services through the Department of the Interior, though communications and power within parks varies widely, with front country buildings often hooked up with fiber or 802.11 WLAN technology and back country buildings supported by contracted satellite services or whatever is available in that terrain.
Within the Park Service, Thornton said “the culture divides ICT pretty strictly from the professional services for A&E project management,” so you have architects, landscape architects, engineers and project managers focused on projects, whereas the full-time ICT people – IT and telecom specialists, radio technicians, etc. – are dedicated to ongoing operations. ICT designers, telecom project managers and ICT infrastructure experts are rare, said Thornton, who was an exception. An outfit called Denver Service Center serves as the central planning, design and construction management project office for the NPS.
The resulting state of ICT infrastructure within the parks is “not in that great a shape overall” because it isn’t done through A&E projects, but rather on an ad hoc basis by in-house electricians and IT or telecom specialists, Thornton said.
“As an ICT professional I look at it and I think this stuff’s like a bumblebee – it shouldn’t fly, but it does,” he said.
As for the Moose project in the title of Thornton’s talk, it began as a rehab job on a single maintenance building, with plans to expand it for use by managers and emergency personnel. But in the late 1990s the project came to a halt when structural problems and asbestos were discovered, and that hiatus lasted until the middle of the next decade, at which point a second value analysis of the project was undertaken (during which a second building on the campus was set for demolition and as a result, the maintenance building would be converted to be the main park headquarters building). The plan was to work on this project over a number of years, due to a lack of immediate funding, but then the American Recovery & Reinvestment Act came along to give hope for a speedier timeline as long as the project was shovel-ready when the funds became available. In the end, the shiny new 53K sq. ft. HQ building was ready in the 2012-13 timeframe, decked out with an impressive standards-based ICT infrastructure that featured a structured cabling system, seven telecom spaces, Cat 6A wiring, VoiP and network security (plus nearby radio communications facilities).
The resurrected building with its snazzy new ICT foundation helped to set the stage for growth in surrounding parts of the park, including several new buildings that used the new outside communications plant and pre-staged conduits.
“But it didn’t come without a whole lot of pain, and a whole lot of fighting, and a whole lot of backlash and finger pointing, and it didn’t come without some really significant unexpected costs,” Thornton said, noting that complications and delays arose well into the construction contract in the 2011 timeframe. He estimates cost overruns
totaled more than $2 million.
Here’s why: despite various analyses undertaken, the team overseeing the project didn’t realize they were going to cut communications across the Moose campus -- and far beyond within the park -- for an indefinite amount of time once the building demolitions began. Those buildings, as it turns out, were distribution points for communications (including VoIP, data, point of sale, SCADA, etc.) to other facilities across the campus and entire park.
Thornton arrived on the scene in 2008, having been previously stationed in Denali, and began inspecting the existing communications plant, which he says was poorly labeled and documented. He wasn’t consulted until deep into the design development phase and by that time Thornton understood enough about the infrastructure to have “the dubious pleasure of telling them you’re about to cause major chaos for your operations when you do this.”
He pointed to a site analysis checklist that failed to catch the basic sorts of things that he saw regarding the ICT infrastructure. He said it wasn’t the A&E firm’s fault, as they did what they were instructed to do, and that didn’t include considering the customer-owned infrastructure as part of an overall utility study. He said the overall value analysis of the project, which had built-in assumptions based on the original project (before the second building demolition was planned) led the team to make the decisions it made despite having numerous opportunities to sniff out problems. And that team, he noted, did not include any ICT experts.
“Unfortunately, this is the normal make-up of a value analysis team in this agency,” Thornton said, adding that the Park Service could stand to take a page from the Department of Defense, which for almost the past 10 years has required such expertise be included in facilities design planning.
Once the Moose project team was made aware of the impending communications disaster, it came up with a fresh plan that included an entirely new outside plant and new 90-foot wireless communications tower (to replace a tower previous on top of the old building).
Thornton wrapped up his talk urging fellow BICSI members to make their voices heard ("We've got a real brand recognition problem..."), as he exhorted them to ensure that customers and contractors on projects have ICT experts on board. A working group might be in order, and Thornton said he's willing to be part of it. Sounds like a perfect retirement project...