ELK GROVE VILLAGE, Ill. – Senior safety officials at United Airlines [UAL] will be meeting at the carrier’s world headquarters here later this month to assess how they might strengthen their safety programs and their safety oversight processes. The goal is to build upon the various pieces of the programs now in place and to produce a more comprehensive and coherent picture of the carrier’s safety status.
“We will be bringing all the safety, security and quality assurance (QA) groups together for the first time since our restructuring,” said Hank Krakowski, UAL vice president for safety, security and QA. Those groups will include flight operations, maintenance, on-board services, ramp operations, and others.
“Our situation has given us an opportunity to really harmonize the whole safety and quality assurance operation,” Krakowski said in a not-so-veiled reference to the carrier’s bankruptcy status. “In the last year, the airline’s operating division has realigned, and our goal will be to assure that the safety functions align with and are optimized under the new structure,” Krakowski said.
Since the terrorist attacks of Sept. 11, 2001 (9/11), the carrier has shed nearly 40,000 employees, shrinking from 105,000 to a diminished but still significant strength of 67,000. In addition, the ranks of upper management have not been spared the turbulence, with some 30 of 40 vice presidents having been replaced since that date.
Out of the trauma of this experience, Krakowski sees potential for progress.
With the second greatest number of flights flown over the past 30 years, United has established a good safety record, but its accident rate – as measured by AirSafe.com – is twice that of Delta Air Lines [DAL], which is the global leader in terms of accumulated flights (see table AIRSAFE). On the other hand, United’s record is considerably better than some other carriers – particularly in recent years. This record includes acts of terrorism, and thus United’s score along with American’s includes the two aircraft each airline lost on 9/11. Its previous safety-related fatal event involved a turbulence encounter on a Dec. 28, 1997 trans-Pacific flight in which one of the 346 passengers was killed and three flight attendants on the B747 sustained serious injuries. Krakowski pointed out that the last crew-caused fatality at United was in the 1978 crash of a DC-8.
A fresh look
Krakowski said the goal of the upcoming meeting is to take a fresh look at operational reliability.
“If we go back to the summer of 2000, operational reliability wasn’t as good as it should have been,” he said. The employee on-the-job injury rate was too high, as was the amount of aircraft damage.
One of the suspect causes, Krakowski said, was a schedule geared to maximize revenue that was “operationally infeasible.” For example, the schedule was oriented to achieve a 20-minute separation between departures and arrivals at the hubs. That brief period left too little time for ground crews to organize for the next flight to be serviced efficiently.
“In the summer of 2002, we went to 30 minutes between departures and arrivals,” Krakowski said. That one adjustment had a significant and positive impact on one measure of safety – ground damage. The amount has trended down some 40 percent.
Other challenges are looming. United plans to outsource more of its maintenance and it intends to extend the heavy maintenance interval for its B757 aircraft from five to six years.
Krakowski said United personnel are on-site at all major airframe and engine outsourcing companies. In addition, he said, “There will be no outsourcing to any firm until my QA people certify the [maintenance] operation independently.”
In a company that has hemorrhaged people, Krakowski said “we are beefing up oversight.”
People are being hired to strengthen QA oversight of both maintenance and operations. “I’ve also added staff for environmental oversight,” Krakowski said.
Data demands analysis
Nonetheless, in an operation as large as United’s, with all of the various programs to track safety, there is a danger of drowning in data. Presently, safety-related information comes from three major sources. The first might be described as data. It consists of operations and reliability reports, outputs from the flight operations quality assurance (FOQA) program, and a host of others, including required reports to the Environmental Protection Agency (EPA) and the Occupational Safety and Health Administration (OSHA).
The second source of information Krakowski described as the “people reports.” He quipped that these reports are known as the “SAP brothers” because of their acronyms. SAP stands for “safety awareness program.” Presently, there are three such confidential employee reporting systems:
- Flight Safety Awareness Program (FSAP).
- Dispatch Safety Awareness Program (DSAP). “A first in the industry,” Krakowski claimed.
- Station Operations Safety Awareness Program (SOSAP). “Soon to begin initial testing,” Krakowski said.
To this triumvirate, a Maintenance Safety Awareness Program (MSAP) is to be designed and implemented, as well as a flight attendant program.
Krakowski said these reports are considered “essential” for the insights they contain. “Airlines that have good self-disclosure systems tend to get more reports,” he said. “We believe we have fewer incidents but our reporting system is better.” In other words, the number of such reports is not a measure of safety, or lack thereof, but rather people’s confidence in the system.
The third major area of safety-related information comes from audits. Within the company, they take two forms: quality control (QC) audits, which are “internal to a division.” For example, check pilots perform a QC function in flight operations, and maintenance inspectors perform a QC function in the maintenance division.
Quality assurance (QA) audits are done by staff outside of the various divisions. The joint quality review teams (JQRTs) are composed of representatives from the airline, the union, and the Federal Aviation Administration (FAA).
Various pieces of this information are reviewed on daily, weekly, monthly, quarterly and, as necessary, ad hoc schedules.
“You want it all to come together at the safety department, and you want management, unions, and others to see it,” Krakowski said.
For example, the FAA is informed of the safety status at United in a weekly conference call involving the FAA’s certificate management office (CMO), the principal operations inspector (POI), principal maintenance inspector (PMI) and the principal avionics inspector (PAI).
“The purpose of this call is to exchange information on what we are seeing,” Krakowksi explained. “This includes an update on the changing structural, financial and operational condition of the airlines, which the CMO can report upstream to the FAA in Washington.”
Some sources of information are considered more useful than others. Regarding the service difficulty report (SDR) system, Krakowksi said, “We haven’t used it much.”
One of the emerging challenges is to align the company’s safety programs with the FAA’s air transport oversight system (ATOS). ATOS is intended to shift the FAA’s traditional emphasis on compliance to more of a process orientation to assure that the carrier’s safety programs are self-correcting.
If the FAA is unable to make the cultural transition from “compliance cop,” resulting conversations with the carrier about ATOS findings, Krakowski said, “won’t be very robust.”
How well are the three major sources of safety-related information working? On a notional scale of 1-10, with 1 being unsatisfactory and 10 being outstanding, Krakowski gave these scores:
- The “SAP brothers”: 8-9. A “very successful” program, Krakowski said.
- Quality assurance: 6-7. “Only because we and the FAA are still adapting from a compliance to a process approach,” Krakowski explained.
- FOQA: 6-7. “There are some good pockets, but much more can be done with FOQA than we are doing, including the final equipping of our entire fleet,” Krakowski said.
And these scores, admittedly judgmental, are nonetheless why Krakowski has called for the meeting of all divisions later this month. His goal is to drive all pieces of the safety program to score in the 8-9 range for “a complete picture that’s reliable and that can be data-mined effectively.” Krakowski’s assessment accords with that the FAA, which found a mix of both shortcomings and strengths during its National Program Review (see ASW, April 1, 2002). Among the principal findings was a “paucity of system analysis.”
From analysis to synthesis
There is a philosophical underpinning to Krakowski’s approach. It is known as “precision from variety.” That is, the more separate pieces of information that are brought to bear, the clearer one’s perception of the situation is likely to be. This doctrine can be illustrated simply. If one looks at a pyramid from the side, it looks like a triangle. Viewed from the bottom, it looks like a square. From the top, it looks like an “X” within a square. All three views combine to give the most precise picture, hence the expression “precision from variety.”
The timing of the upcoming meeting is significant. United announced late last week a 10 percent increase in the number of daily flights in June, up to 1,722 daily flights from 1,560 daily flights presently. As that level of flying activity increases, Krakowski is looking for increased assessment of the overall safety situation.
An Independent Look at United’s Safety Programs
Highlights of the FAA’s National Program Review:
Continuing Analysis & Safety Surveillance System (CASS):
Strength: Auditing function clearly separated from QC.
Weaknesses: Procedures are not standardized. Little overall system analysis.
Reliability program:
Program focuses on those issues with the highest negative impact on safety and reliability. Self-imposed hard times on components to improve overall reliability.
Safety program:
QA department has a database for human reliability as well as for trend analysis. The team found few concerns at United.
Source: FAA
Three Aspects of Safety Awareness at United
To pilots:
“Every day UA operates with hazardous materials (HAZMAT) in the cockpit. This HAZMAT is in the form of consumable liquids (coffee, soft drinks, milk, orange juice, drinking water, etc.). This HAZMAT causes inflight … ‘irregular’ actions on a recurring basis. Several UA incidents involving cockpit spills have resulted in emergencies, at least one in a cockpit electrical fire, and several diversions and/or evacuations.
“The areas of the airplane that receive the greatest danger are the center pedestal, throttle quadrant, and area just forward of the throttle quadrant. Significant damage, however, also occurs in cockpit flooring, systems components (TCAS, air data computer, flight management computer, autopilot computers, yaw damper couplers, etc.) in the E&E bay below the cockpit floor, nose gear components, and oxygen masks.
“The current cups, cans and bottle used in the cockpit have insufficient spill protection. Probably the most effective container would be a cup with a trigger-controlled lid.”
– From United’s Safety Liner (the company’s safety journal), Winter, 2003
To flight attendants:
“Flight attendants should be cautious when putting their hands into an airplane’s seat pockets. Recently, a passenger blindly put his hand into a seat pocket and was stuck by a syringe that was not his. Be careful – always look before putting your hands into the seat pocket.”
– From, Partners in Safety (United’s publication for flight attendants), Spring 2003
To maintainers:
“The root cause of the [engine failure on takeoff] was determined to be 3 missing bolts that attach bridge connectors to the 3rd stage variable stator vane (VSV) unison ring.
“Without the bolts holding the bridge piece to the ring, the low ring rotation was solely controlled by the upper ring rotation. Over time, friction increased with the bridge and the upper ring developing radial movement … [which] allowed the unsecured bridge to make contact with the actuating arm like a wedge to pull the arm from the ring.
“The now mislocated lever caused one of the VSV to be out of position. This caused the resultant airflow to impact the 4th stage rotor blades like a pneumatic hammer, eventually causing the blade to fail. The damaged blades resulted in developing a titanium (Ti) fire in the engine.”
– From United’s Safety Liner, Winter, 2003
|
Fatal Event Rates for Selected Carriers Since 1970
|
|||||
|
For carriers with 4 million or more flights
|
|||||
|
Carrier
|
Rate
|
Flights
|
FLE*
|
Events**
|
Last
|
| Delta Air Lines/Delta Connection |
0.16
|
20.0 M
|
3.24
|
6
|
1997
|
| Continental Airlines/Cont. Express |
0.18
|
8.0 M
|
1.47
|
5
|
1997
|
| Lufthansa/ Condor |
0.19
|
7.3 M
|
1.41
|
3
|
1993
|
| SAS |
0.19
|
5.4 M
|
1.00
|
1
|
2001
|
| British Airways |
0.22
|
6.35 M
|
1.40
|
2
|
1985
|
| Northwest Airlines/ NW Airlink |
0.28
|
9.2 M
|
2.61
|
4
|
1993
|
| United Airlines/ United Express Express |
0.35
|
14.4 M
|
4.97
|
9
|
2003
|
| United Airlines/ United Express |
0.37
|
18.0 M
|
6.69
|
11
|
2001
|
| American Airlines/ Am. Eagle |
0.54
|
17.0 M
|
9.23
|
12
|
2001
|
| Air France/ AF Europe |
0.55
|
5.9 M
|
3.23
|
7
|
2000
|
|
For selected carriers with less than 4 million flights
|
|||||
| Swiss/ Crossair |
1.20
|
3.20 M
|
3.83
|
5
|
2001
|
| Singapore Airlines/ SilkAir |
1.50
|
1.00 M
|
1.50
|
2
|
2000
|
| Korean Air |
2.58
|
1.30 M
|
3.35
|
7
|
1997
|
| Turkish Airlines |
6.83
|
1.10 M
|
7.51
|
9
|
2003
|
| China Airlines (Taiwan) |
7.16
|
0.90 M
|
6.44
|
10
|
2002
|
|
* FLE: Full loss equivalent. The sum of the proportions of passengers killed for each fatal event. For example, 50 out of 100 is an FLE of 0.50, 1 of 100 would be an FLE of 0.01. The rate is computed as the FLE divided by the number of flights in millions. |
|||||