Aeronautics Study - Safety Threats

Aeronautics Study - Safety Threats

Six near runway collisions at San Francisco, New York, Ft. Lauderdale and other airports were narrowly averted in just the last six months leading the NTSB to highlight the issue of runway safety as among its most important issue areas to be addressed by the Federal Aviation Administration (FAA)." (NTSB, 2007) This disturbing lead-in introduces the release the National Transportation Safety Board (NTSB) issued November 8, 2007. (Appendix a) During its public meeting which focused on needed critical changes federal agencies need to implement to reduce accidents (including runway incursions) and save lives, the "NTSB reviewed its 'Most Wanted List' of safety improvements," (NTSB, 2007) established in 1990.

The unsafe behavior of human operators is one of the most pressing threats to the safety of complex technological systems. It has been estimated that human error is involved in 58% of medical misadventures (Leape et al., 1991), 70% of aircraft accidents (Hawkins, 1993), and 80% of shipping accidents (Lucas, 1997). Maintenance error is becoming a significant area of concern to the aviation industry. Of 14 major accidents recently investigated by the National Transportation Safety Board, 7 involved maintenance deficiencies ("Human Factors Programs," 2002). In addition to the human costs of accidents, maintenance errors impose a significant financial burden on airlines, as they are a major cause of flight delays and cancellations (Marx & Graebel; 1994). (Hobbs & Williamson, 2003)

Figure 1: Headed for an Incursion?

A runway incursion is "Any occurrence at an aerodrome involving the incorrect presence of an aircraft, vehicle or person on the protected area of a surface designated for the landing and take off of aircraft."

2007, October 10). http://www.faa.gov/runwaysafety/

Pilot Error or.

Kirwan, Rodgers and Sch fer (2005, p. 44) report that within pilot deviations, smaller aviation aircraft likely to be involved in a runway incursion is less than commercial aircraft. These authors contend the following areas merit consideration for change:

Procedures

Training

Technological Enhancements

VPDs: Vehicle and Pedestrian Deviations

VPDs).

In "DWI Convictions Linked to a Higher Risk of Alcohol-Related Aircraft Accidents," a diurnal article by Kathleen L. Mcfadden (2002) data presented show "alcohol to be a relatively minor problem in general aviation, accounting for less than 1% of the total number of accidents."

One Study

METHODS

The data consist of the records of 380-912 pilots who flew in general aviation and obtained a Class III Medical Certificate during the 10-year period from 1988 to 1997. The data were obtained from the FAA's Consolidated Airman Information System (CAIS). The CAIS database contains information on DWI convictions that have been verified by querying the National Driver Register as well as pilots' demographic information (e.g., age, gender, recent flying hours).

The accident data were obtained from the FAA's Aviation Standards Accident/Incident Database (AID). The AID consists of specific information about accidents, including the accident date, accident time, aircraft make, aircraft model, cause factors, and fatalities. Accident data also span a 10-year period (1988-1997) and include only those cases in which the NTSB determined that alcohol was a contributing factor in the accident. It is possible that during the period of this study the FAA data could have been contaminated because of a lack of control for postmortem alcohol production. However, the more conservative NTSB guidelines are used here, so it is unlikely that any postmortem alcohol production cases were included in this study. In order to complete this study, the two databases (CAIS and AID) were subsequently merged. Alcohol-related accidents involving noncertified pilots were not included.

One limitation of the CAIS database is that it does not include the exact date of DWI convictions. Although the FAA keeps hard-copy files of this information, I was allowed access only to the databases. However, prior in-depth investigations conducted by the FAA revealed that in more than 70% of the cases in which I found that pilots had DWI convictions, the DWI preceded the accident (McFadden, 1997a; McFadden & Hosmane, 2001).

A logistic regression model was developed to determine whether pilots with DWI convictions were more likely to have alcohol-related general aviation accidents than were those without convictions. The dependent variable was whether or not the pilot had an alcohol-related aviation accident. Logistic regression is considered the preferred method for handling binary responses (Hosmer & Lemeshow, 1989). It is one of the most widely used functions for modeling the probability of an event. This is because the functional relationship between the probability of an event and the independent variable is approximately sigmoidal, or S. shaped. The logistic function f (x) is also S. shaped, so it approximates this relationship. It also guarantees that the probability will range between 0 and 1, inclusively. The ease of interpretability of the model parameters is yet another benefit of logistic regression.

The primary predictor variable

Prior studies on the effect of age on accident rates have reported conflicting results. Airline pilots are less likely to have pilot-error accidents and incidents as their age increases (McFadden, 1996, 1997a, 1997b, 1998). However, age is not typically associated with accident rates among general aviation pilots (Li, 1994).

Three of the prior studies on airline pilots (McFadden, 1996, 1997a, 1997b) include data on general aviation accidents, so they are especially relevant to the current study.

Another factor believed to affect accident rates is recent flying hours (a pilot's flying hours in the last six months). Recent flying hours is a measure of the level of exposure of pilots to risk and a measure of pilots' currency (Li, 1994). The theory is that pilots who fly more frequently may be exposed to a greater risk of being involved in aviation accidents (McFadden, 1996). Consequently, if one pilot had more current flight time, this may explain his or her higher accident rate.

Primary Predictors

Human Factors Impacts in Air Traffic Management

By Barry Kirwan, Mark Rodgers, Dirk Sch fer

In research and application of Human Factors in Air Traffic Management (ATM) systems design, development and operation, there remains a lack of clarity. This book seeks to redress this situation by presenting case studies of human factors applications in which there is demonstrable success in terms of improvement in operational systems.

Kirwan, Rodgers and Sch fer (2005)

Within pilot deviations, we find that smaller general aviation aircraft are more likely to be involved in a runway incursion and commercial aircraft. A study of runway incursions from 1998 through 2001 revealed a rate of five runway incursions permeate operations, i.e. takeoffs and landings (FAA, 2002). While general aviation aircraft comprise approximately 58% of the total operations, they are involved in approximately 65% of the reported runway incursions (FAA, 2002). Il, there is an average of one serious than (a or B) runway incursion involving two jets somewhere in the national aerospace is system (NAS) each month.

Several studies have been conducted to examine the causes of these incidents and to identify solutions to the underlying problems. The studies provide a strong foundation for the identification of pilot, controller, and environmental factors that contribute to runway incursions. They also identify a potential remedies that include recommended changes to procedures, training, and technological enhancements.

Error Definition

Perceptual error a failure to detect a sign that the person was attempting to detect

Memory lapse the omission of an action that the person intended to perform

Slip the performance of a familiar skill-based action at a time when this action was not intended, or the failure to carry out such an action correctly; this category included fumbles and trips

Rule-based error a failure to correctly invoke familiar rules or procedures, either written or based on experience, when dealing with routine problems or when making decisions in familiar situations

Violation an intentional deviation from procedures or good practice

Knowledge-based error an error in a situation that was unfamiliar or that presented new problems for the person, for which neither automatic mappings nor rules existed

Mischance the person adhered to correct procedures, but his or her behavior was nevertheless instrumental in leading to the occurrence

TABLE 2: Contributing Factor Taxonomy Used in the Current Study

Factor Definition

Fatigue Mental or physical fatigue, generally related to a lack of adequate nighttime sleep and/or to night shift work

Pressure Work was being performed under unusual time pressure or haste

Coordination Inadequate teamwork and communication between workers

Training Factors relating to inadequate training of personnel

Supervision Factors relating to inadequate charge of workers

Previous deviation Incorrect performance of a task at an earlier time, and this error remained latent and was not recorded as an event in the occurrence sequence

Procedures Poorly designed, poorly documented, or nonexistent procedures, or when a deviation from procedures was routinely accepted by management and/or operational personnel

Equipment Includes poorly designed or maintained equipment or tools, or a lack of necessary equipment, including aircraft spare parts

Environment the physical environment in which the work was being performed, which was beyond the control of the worker -- for example, darkness, glare, heights, and excessive noise

Physiological the worker's performance was affected by a medical condition or by sensory or physiological limitations

TABLE 3: Occurrence Outcomes

Outcome N. Percentage (a)

System operated unsafely during maintenance 80 13

Incomplete installation 48 8

Maintenance worker contacted hazard 45 7

Incorrect assembly or location 44 7

Towing event 44 7

Vehicle or equipment contacted aircraft 31 5

Material left in aircraft 27 4

Wrong equipment or part installed 23 4

Part not installed 22 4

Part damaged during repair 21 3

Panel or system not closed 21 3

Required service not performed 20 3

Failure of component or tool 15 2

Fault not found 15 2

Falls and trips 14 2

System not made safe before maintenance 12 2

System not reactivated 10 2

Pin or tie left in place 9 1

Documentation error 9 1

Note. For an additional 14 occurrences, the outcome could not be determined.

A a) Figures are rounded to nearest percentage.

Errors

Perceptual error

Example: After being on duty for 18 hr on a long overtime shift, the worker was carrying out a general inspection on an engine at around 22:00. He missed obvious damage to the internals of the cold stream duct area. The damage was found later, when another defect was being investigated.

Memory lapse

Example: Just prior to the departure of the aircraft, I remembered I had left a blanking plug within the engine inlet area. I advised the pilot that I needed to check that area again and retrieved the blank.

Slip

Example: Without thinking, I moved to wipe oil with a rag. The rag was ingested in the engine intake causing FOD [Foreign Object Damage].

Rule-based error

Example: A mechanic did not check the position of the flap lever before he pushed in a cockpit circuit breaker that provided electrical power to a hydraulic pump. When the pump started, the flaps began to retract automatically. This could have caused damage to the aircraft or injured other workers.

Violation

Example: At the end of a shift we realized that an engine hadn't been run to check for oil leaks when the aircraft was to be placed online. Under pressure to avoid a delay due to this oversight, the run was carried out too quickly and the engine was not un-cowled properly to check for oil leaks and consequently after departure that particular engine ran out of oil as the result of a damaged seal.

Knowledge-based error

Example: I wanted to turn the radio master on but could not find it, as the switches were poorly marked or unreadable. I was unfamiliar with the aircraft, so I asked an airframe tradesman who was working on the aircraft and he pointed to a red rocker switch. I queried him and he said that must be it. I pushed the switch and the right engine turned over, with the propeller narrowly missing a tradesman who was inspecting the engine. There is no radio master in this aircraft. I immediately marked the "start" and some other switches and learned a valuable lesson.

Mischance

Example: A service procedure was carried out in accordance with the aircraft maintenance manual. The manual however, contained an error, which resulted in an aircraft system failing to operate correctly during a functional test at the end of the maintenance procedure.

Factor n or Fatigue 1-0.2

Pressure 8-1.9

Coordination 6-1.4

Training 1-0.2

Supervision 5-1.9

Procedure 16 9.0 **

Equipment 1-0.3 deviation 4-4.3 *

Environment 1-0.8 chi square](9, N = 805) 46.68

Wald test significant at p

Memory lapse

Violation

Unclassifiable behavior

Knowledge-based error

Slip

Rule-based error

Hardware event

Perceptual error

Mischance

Environment event

Figure 4. Factors contributing to maintenance occurrences.

Percentage of occurrences involving event

Pressure

Equipment

Training

Coordination

Fatigue

Procedure

Supervision

Environment

Previous deviation

REF

Runway Safety

The Human Element

Runway safety is one of the FAA's highest priorities - specifically the problem of runway incursions. A runway incursion is defined as: Any occurrence at an airport involving an aircraft, vehicle, or person on the ground that creates a collision hazard or results in a loss of separation with an aircraft taking off, intending to take off, landing, or intending to land.

Though relatively few in number when compared to the massive amount of traffic that moves safely through our nation's airports every day, runway incursions present a special problem. Not only do they have the potential to put more lives at risk due to the number and proximity of aircraft operating on the airport surface, they also take place in a complex and dynamic environment where root causes are difficult to isolate.

At the simplest level, incursions occur because people make mistakes. Humans are superbly skilled at making decisions under a wide range of circumstances but, for a variety of reasons, they are also fallible. Consider this human vulnerability within the context of the numerous variables that may contribute to human error and you can appreciate the problem. Its not just a pilot, controller, or vehicle operator problem, it a problem that all of us in the aviation community share.