Monday, September 1, 2008
Fleet Management: Just the Facts
Spurred by big operators, and some small ones, manufacturers are pursuing new tools to reduce maintenance burdens and costs for helicopter fleets.
AOG is a term that raises temperatures and blood pressures. The abbreviation for aircraft on ground implies interrupted service and lost revenue for an operator and challenges to the reliability of and support for a manufacturer’s products.
"A lot of folks would run around when there was an AOG, all trying to do the right thing and get the right part, and everybody’s hair would be on fire," said Joe Eltman, manager of life cycle analysis and planning for Sikorsky Aerospace Services.
But AOG can also be a ploy for an operator — to jump the line waiting for a part in short supply or to shift the blame for aircraft failures caused by deferred maintenance rather than reliability shortfalls.
AOG is just one of many aspects of helicopter operations that often have been governed more by emotions than facts. That may be changing, though, if a number of industry initiatives bear the fruit their proponents intend.
Manufacturers like Sikorsky Aircraft and Turbomeca (see story on page 40) have separate, major efforts under way to track in near real-time the performance and configuration of their products. Eventually, "we will be able to customize the time before overhaul (TBO) of each engine" based on the data collected on its utilization and maintenance, said Serge Maille, the Turbomeca official in charge of that engine maker’s effort.
Operators are pushing for more accurate measures of aircraft reliability as a means of slashing the cost of keeping helicopters in the air. Leading the charge for what is called condition-based maintenance is the world’s largest operator of rotorcraft, the U.S. Army. For example, it last year changed the name of its main aviation command to the Army’s Aviation and Missile Life-Cycle Command, adding "life-cycle" to reflect the new focus.
Even some small operators are pushing for tools to better track and improve the performance of their aircraft. Helicopters (NZ) Ltd in Nayland, New Zealand is constantly pestering Bell Helicopters to tweak its maintenance and support programs for the Model 412s it flies, according to Bell officials. The operator’s insistence is based on data collected by the Intelligent Automation Corp health-and-usage-monitoring systems (HUMS) Bell placed on the 412s. Drawn in part by the projected demand for greater data-collection and monitoring services from operators and aircraft makers, Honeywell last month acquired Poway, Calif-based Intelligent Automation.
A leading effort to meet that anticipated demand is Sikorsky’s, which aims to build on the recording and monitoring elements integral to its S-92 in creating a predictive-analysis capability for all of its aircraft products. The nerve center of that effort is the manufacturer’s Fleet Management Operations Center.
As Eltman explained it, Sikorsky President Jeff Pino "is really driving the entire company towards improving the reliability of our components. That is the most important thing. If we do that, everything else follows."
Eltman is one of the core team charged several years ago with setting up the Fleet Management Operations center to track the usage and analyze the reliability of the S-92 from the first days it entered service. Housed within the headquarters of Helicopter Support, Inc (HSI), Sikorsky’s parts and service subsidiary in Trumbull, Conn., the center began work in April 2005.
The center’s objective is to better understand how Sikorsky aircraft are operated and maintained on a daily basis and why their performance and reliability varies, for better or worse, from projections engineers developed in designing them. The data its staff collects and crunches is intended to enable managers of each Sikorsky aircraft program to better focus their efforts in improving reliability and decreasing direct maintenance costs.
The sources the center draws on include each customer’s organic maintenance management system and data from the Goodrich HUMS units installed on S-92s. HUMS data, transmitted each night to a Sikorsky server, provides operational information on each aircraft. The center also taps into the supply data system of HSI, which is one reason why it is co-located with HSI in Trumbull.
"A lot of the benefits the customer is seeing HSI is seeing as well," said Kelly Dixon, director of flight management solutions at HSI and a key player in the effort. "The better we can predict the demand [for parts], the more cost it saves HSI and the more parts are available for customers."
When the center’s staff first began briefing top Sikorsky management on the S-92’s performance, Pino praised the data they’d collected but noted they didn’t have the power to act on the trends revealed by it. The aircraft program managers did, so Pino charged them with presenting the collected data to top managers — and plans for addressing adverse trends. That last part is key.
"We don’t just do data analysis for data analysis’ sake," Eltman said. "We actively use it to drive the business and help the customer."
While the center is focused on the S-92, it is intended as a proof of concept for all of Sikorsky’s aircraft. The S-92 effort is the most in depth. Legacy UH-60 Black Hawks, for instance, don’t have HUMS. Newer UH-60Ms do, and some A and L models have been retrofiited with Goodrich HUMS. The new S-76D will have a HUMS, but such systems are a customer preference on earlier model -76s. "When we started this process," Eltman said, "we wanted to make sure whatever process we put in place was transferable to the whole enterprise."
The overall goal of that process is fourfold: improve the design of Sikorsky products to make them more reliable and easier to support through the life of an aircraft; identify and fix problems that impede customer support or drive the cost of it up; lay out the requirements of logistics support throughout an aircraft’s life; and develop a single database for providing logistics support for Sikorsky’s fielded aircraft, production lines and spares inventory.
Center staffers concentrate on understanding what factors are driving S-92 maintenance, what impact they have on customers and what can be done to mitigate that impact. They use a United Technologies Corp method called Russo charting for ranking components based on what is important to the aircraft operator and the aircraft program, combining that with data on direct maintenance costs. Each Sikorsky program is challenged with reducing the direct maintenance costs for its aircraft by 10 percent each year.
But fulfilling those goals takes more than crunching numbers. It requires the insight and cooperation of S-92 operators: insight into why they operate and maintain their aircraft in unique ways and cooperation in bearing the cost of the data-recording and monitoring equipment and the services Sikorsky sells based on that gear.
The operator payoffs Sikorsky envisions are numerous: a reliable supply of parts when they are needed, greater reliability from those parts, more predictability in the maintenance and failure of parts, and a significant reduction in inspection requirements (see "Target: 40 Percent Cut in Inspections," page 38).
An example is the S-92 tail-rotor blade yoke. Sikorsky found demand for replacement parts was much higher than projected, which left customers waiting on that part. In crunching fleet maintenance data, center staffers discovered that about half of unscheduled removals of the yokes coincided with a 1,250-hr inspection of an operator’s S-92. Some operators were pulling the yokes whether they needed replacement or not. They figured it was less disruptive to do so at a major check than to risk an unscheduled removal of the yokes later. But Sikorsky hadn’t based its supply of replacement yokes on that. "We were able to look at that data from a customer-specific behavior perspective," Dixon said. "By balancing that against the scheduled maintenance calendar based on the center’s projections, we were able to come up with unscheduled demand predictions aligned to 1,250s. The power of that is huge."
Using that data, center staffers and HSI personnel in April came up with a 12-month forecast of yoke demand, "and we’re seeing that come to life," Dixon said. "Where this part was always a shortage in terms of customers needing this item, it’s actually on the shelves, stocked and allocated where it needs to be. We’re looking to transfer that to the rest of the big maintenance drivers."
Achieving those goals also takes a mix of specialists in reliability and maintenance engineering, maintenance practices and the logistics to support a helicopter operation. That mix is reflected in the center’s staff, which has reliability and maintenance engineers working side by side with logisticians like Eltman, a former U.S. Marine Corps helicopter mechanic. Those people have contrasting, at times conflicting, cultures.
Engineers are used to designing and producing aircraft through rigorous analysis. "Logisticians like me are used to flying by the seat of their pants and making estimates on the fly," Eltman said. Where some of the conflict lies, he said, "is trying to come to that middle ground that says, ‘Look, I want to put a little bit more rigor into the process, but I don’t want to do an engineering analysis.’ So that’s been a real learning experience for everybody in the room."
In truth, the efforts embodied in the Fleet Management Operations Center are all about culture change — within Sikorsky, within its base of customers and within the industry as a whole.
Like many a manufacturer, Sikorsky prospered for years not through the sale of aircraft but the parts and services to keep them flying. That held little incentive to improve product reliability.
Adoption of "performance-based logistics" contracts has fostered efforts to improve reliability, but only in a limited way. Sikorsky, for instance, has such a contract with the U.S. Navy that called for it to deliver 85 percent of requested parts within 48 hr of their request. At one point, Sikorsky was delivering 90 percent within that time frame. That prompted a top executive to question why the company was exceeding the contract requirements if there was no additional money to be made by doing so. There was no point in improving the reliability of parts or their supply, so the thinking went, unless that boosted profits.
That has made perfect sense. But helicopters have become a commodity, with those in the same class carrying roughly the same loads over similar ranges at like speeds. The key discriminator is the cost of doing so. Since maintenance is the largest cost component for a helicopter, reliability is becoming the competitive battleground.
The specialists in the center spend a good deal of time analyzing signatures of sensors placed on critical dynamic components. They compare those from individual aircraft to those from the fleet for signs of a problem.
The approach seems to work. In one case, vibrations showed up in one aircraft’s swashplate. They were well within design limits, but higher than for any other S-92 flying. Sikorsky alerted the operator, who had found no other indications of a problem. The center staff continued to monitor the vibrations, and Sikorsky pre-positioned a new swashplate. They removed the swashplate and found its bearing cage was cracked.
In another case, an oil cooler was changed out. Prior to the change, its aircraft was showing trends well within the fleetwide numbers. Afterwards, its numbers were running higher than the fleet. In examining the new cooler, mechanics found a seal had been crushed. Investigating it further, Sikorsky found the seal was crushed by a vacuum packing process used by the cooler’s manufacturers. That enabled Sikorsky to have the process fixed, pull damaged coolers from its inventory and recall those delivered to customers, averting maintenance problems downstream.
As data and analyses accumulate with time, the Fleet Management Operations Center is becoming more sophisticated. A case in point is a failure of a tail pitch-change shaft on an S-92 whose onset went undetected by the aircraft’s monitoring systems. In reviewing the failure, engineers found the imminent failure went undetected because — due to configuration constraints — a decision had been made not to put a sensor on that shaft. But, as Eltman tells it, the engineers dug a bit further and found that by examining and comparing historical readings from several sensors near the shaft, they could see signs of the failure. Those findings were fed back into the programming of the Goodrich HUMS on the S-92 so that mechanics can run a test for a tail pitch-change shaft failure with a computer program instead of a physical inspection.
But those and future results of the center’s work rely on sensors. Mechanics and pilots don’t know the sensors. They do know the sound and feel of their aircraft systems, and trust their interpretations of them.
The technology to use sensors to monitor critical systems is proven, but faith in them is not widespread. The success in the efforts of Sikorsky and others requires a leap in faith.
The promise of those efforts lies in "being able to analytically change component lives and analytically tell folks to leave parts on the aircraft because they’re different but they’re okay," said Eltman. The center’s data "at least brings the facts to the table to allow you to get there."
Target: Near-40 Percent Cut in Inspections
Sikorsky is using the predictive-analysis capability being developed through its Fleet Management Operations Center to overhaul regimes for maintenance and inspection of its S-92s.
Center staff speak of the effort as an attack on the old regimes, which are based largely on conservative, hard-time limits on the lives of critical components. Those limits traditionally have given little consideration to how components are actually used in the field or how differences between, say, a helicopter flying executive transport and one doing firefighting missions might extend or shorten component life projections. That is because engineers lacked the data needed to quantify those differences — until now.
With the S-92 fleet fitted with health and usage monitoring systems (HUMS) and having surpassed 100,000 flight hours, the engineers and logisticians in the Fleet Management Operations Center have enough data to begin their attack, which is three-pronged.
First, they are working with aircraft program managers and operators to scour the S-92 maintenance program for hard-time inspections that were based on limited or no criteria. That is, engineers had no good data to project how a component would perform in the field, so they imposed inspection requirements based on the most conservative estimates of when that part would fail.
The second prong targets existing inspections that are ineffective, in the sense that they have failed during the S-92’s time in service to detect indicators of component failure.
Lastly, they intend to eliminate inspections of components fitted with HUMS sensors. A challenge there is the need to convince maintainers in the field that they don’t need to inspect a component because it is essentially monitoring itself.
"We think we can take almost 40 percent of scheduled inspections out of the S-92 maintenance program" through this campaign, Eltman said.