Maintenance mistakes/faults influence the safety of airplanes and have so a hugh failure potential. Up to 30% of the inflight shut downs (IFSD) and 15 % of flight accidents of airliners trace back to maintenance mistakes at airlines (Lit. 19.2.5-3). At every human fault which lead to a serious incident, there will be by guess 600 further however not so serious mistakes.
The definitions of terms like errors and violations contains Ill. 19.2.5-1 (Lit. 19.2.5-5). A statistic of the most frequent causes shows Ill. 19.2.5-2.
To avoid failures a preventive analysis can help. With this maintenance procedures respectively its sequences are investigated for weak points. Are these identified measures will follow to exclude them. To this kind of weak point analysis belongs also the so called FMEA („failure mode and effects analysis“, volume 4, Ill. 17.1-10). To determine the cause also serves a so called problem analysis (PA, volume 4, Ill. 17.1-11). Could this be realised, there is a good chance of a successful targeted remedy. A PA is applicable at technical and nontechnical problems and allows a comprehensible documentation. In case the failure cause respectively the technical reason of the problem in the maintenance is cleared, a special form of the problem analysis was developed and introduced, the so called „maintenance error/event decision aid” (MEDA). The particularity of this procedure is to determine contributory factors on maintenenace errors. For this it is of crucial importance to find concerned persons for questioning. It has shown, that about 80 - 90 % of the contributing factors must be seen in connection with the management. For only 10-20% the technician or the inspector is responsible. Are the contributing factors known, the precondition for sustainable remedies is established, especially for weak points in the organisation respectively the management.
This by the boeing company for the aviation developed procedure uses forms/sheets to guarantee the right sequence and the completeness (Ill. 19.2.5-2, Lit. 19.2.5-4). The forms can be purchased in the internet (see Lit. 19.2.5-4). In it the field of human factors (chapter 19.1) plays a special role. Because the satisfying answer of the forms is demanding the Boeing company has created a guidance („users guide“, Lit. 19.2.5-4), which also can be obtained by using the internet.
Illustration 19.2.5-1 (Lit 19.2.5-5): If we are engaged with problem- analysis and failure analysis the definition of the usual terms is a requirement. In the specialist literature we can find schedules in different „mistake variants”. They deal with mistakes/errors by
In MEDA in contrast targeted descriptions of the different mistakes are used (examples in the upper frame). There are also indicated after evaluation of the ICAO causes for mistakes equivalent to its likelihood.
To avoid a confusion in MEDA, errors are distinguished from violations (lower frame). Typical violations are highlighted in examples. The US Navy uses in an own MEDA like process (human factors analysis and classification system = HFACS) three main types of violations: Routine, infraction and exception. The United Kingdom Flight Safety Committee allocates ten causes under consideration of the likelihood in a so called mandatory occurrence report to so called maintenance mishaps (Ill. 19.2.5-2).
Illustration 19.2.5-2 (Lit 19.2.5-5): Result of of a systematic analysis of incidents by the United Kingdom Flight Safety Committee. Violations are bold type marked because of their especial safety relevance.
Causes for such „mishaps“ are compiled in the so called „dirty dozen” (Ill. 19.1-1.2).
Illustration 19.2.5-3 (Lit 19.2.5-5): A contributing influence promotes errors and violations. In the technical English literature we speak a little more circumstantial of as „performance shaping factor“. Under a contributing factor everything will be understood, what influences the work of an technician or a mantenance inspector. To this belongs a poor illumination of the work place (Ill. 19.1-8), incorreect tools/devices and the disruption or the process (Ill. 19.1.1-7 and Ill. 19.2-13). Not so striking are unclear informations from supervisors (Ill. 19.1.1-1), management decisions dating back for a longer time or the certification of the work of a not approved technician (Ill. 19.2-2.2). The chart applies to a technician in an usual maintenance environment and surveillance. The different frames stand for organisational levels which can act as contributing factors at his work.
The guideline in the specified literature gives hints, how those factors/influences in the completed MEDA forms will affect.
Illustration 19.2.5-4 (Lit 19.2.5-4): MEDA is a form sheet supported investigation of maintenance failures. An especial attention applies to the determination of contributory influences/factors. These have shown as especially important and are a key to effective remedies.
Forms (sheets) can be purchased by the internet. Also the „maintenance error decision aid (MEDA) users Guide” (Lit. 19.2.5-5), which explains the procedure. In the foreground stands to a lesser extent the question after the causer than „why the problem occurred“.
The MEDA process is exactly ruled in all its single steps (scheme above left and Ill. 19.2.5-5).
The forms/ sheets guide the user through those steps whose questions are allocated five parts. They are shortly displayed in the frames.
Illustration 19.2.5-5 (Lit 19.2.5-5): In the following the single steps of a MEDA analysis for the determination and classification (error/mistake or violation, Ill. 19.2.5-1) of the contributing influences are considered.
„1 and 2” Occurring of an incident: First with a (systematic) problem analysis (volume 4, Ill. 17.1-11) the technical causes of the problem/failure, its mechanisms and sequences will be determined. If the problem analysis shows a connection with a maintenence process carried out by a technician or inspector, MEDA can be reasonable introduced.
„3“ In contrast to the solely technical oriented problem analysis it is in case of a MEDA of crucial importance to find the concerned persons (technician, inspector) to question them.
„4” Questioning of the concerned maintenance personnel. There are two target courses:
This task is very demanding. Therefore helpful forms/sheets (Lit. 19.2.5-4) and an additional guidance (Lit. 19.2.5-5) have been developed. With own improvement proposals the questioned can be, different as in a causer role, motivated involved in the process.
„5“ Correspondent with the questioning result in „4” the questioning can be expanded at a further field like the planner of the maintenance processes, responsible persons for the work environment (e.g., light, tools) or the spare part supply.
„6“ After finshing the questionings, the results/filled in forms read into a data bank. Here an analysis for possible trends during incidents, errors, violations and contributing effects takes place. The effectivity of such an analysis depends from the as high as possible number (>20) of already existing data of other cases.
„7” Assessment and decision to introduce recommended improvements. Such proposals to avoid contributing influences can be extensive (time, costs). They must be approved/concluded by the management. Possibly before a decision further investigations must be carried out. Suitable is a `decision analysis' (=DA) and an `analysis of potential problems' (= APP, see volume 4, Ill. 17.1-11)
„8“ Feedback of the remedies/improvements to the concerned maintenance personnel. Not only the information content is of high importance for the success of the remedies. Essential is also the motivation of the team and the shown demonstration of the target course for improvements and not in direction of a punishment.
Illustration 19.2.5-6 (Lit. 19.2.5-5): In its must simple form MEDA consists of two above shown basic models. In the first case (upper basic model) from a contributing factor (CF) develops an error and from this the incident. In cases like a forgotten O-ring seal of a magnetic plug in the oil circuit (Ill. 19.2-2.1/-2.2 and Ill. 19.2-3.1/-3.2) it comes certainly to an oil leakage during take off power. In the most cases the likelihood of the causality between the steps may not be near 100 %. For this reason there will be likelihoods introduced between two steps. Frequently not only one contributing factor exists. So such an factor can act on a further one and so on. For example lacks of clarity in manuals and instructions can intensify information problems during shift handover and these can be aggravated by organisational weak points (Ill. 19.1.1-6.1). Also several independent chains of contributing factors can affect the error and in the end the incident.
Is there a violation, that means a intentional aberration from specifications, this can directly influence the incident (lower basic model). But this can also happen indirect by triggering an error.
In the lower frame connections and sequences for a complex, however quite realistic case are demonstrated.
19.2.5-1 G.J.Fogarty, R.Saunders, R.Collyer, „Developing a Model to Predict Aircraft Maintenance Performance”, Beitrag zum „10th International Symposium on Aviation Psychology“, May 3rd - 6th, ColuSmbus Ohio, page 1-6.
19.2.5-2 E.G.Tripp, „Human Factors in Maintenance”, Zeitschrift „Business&Commercial Airlines“, April 1999, page 82-87.
19.2.5-3 V.P.McConnell, „Workforce Issues: Essential Human Factors”, Zeitschrift „Defense Daily Network“, www.aviationtoday.com, Mar 24, 2006, page 1-6.
19.2.5-4 „Maintenance Error Decision Aid (MEDA) Results Form”, www.chirp-mems.co.uk, March 24, 2006, page 1-6.
19.2.5-5 „Maintenance Error Decision Aid (MEDA) Users Guide“, Fa. Boeing, , 2006, page 1-51.
19.2.5-6 R.Hughes, Interview in der Zeitschrift „Aviation Today”, www.aviationtoday.com, Mar 24, 2006, page 1-3.