Safeskies 2003 - Maintaining the Focus: The ATSB Perspective

[slide 1] Note: Presentation slide references.

Thank you, Mr Chairman and good morning ladies and gentlemen.

As most of you know, the Australian Transport Safety Bureau was established on the 1st of July 1999 and you will see in your delegate's information folders that the ATSB's mission [slide 2] is to maintain and improve transport safety and public confidence through independent 'no blame' investigations which seek to uncover the causal factors that led to an accident or incident and encourage safety action to prevent accidents in future. We do this in aviation, marine, and now interstate rail [slide 3].

In addition to occurrence investigation, the ATSB is building its aviation safety data analysis and research capacity which may also lead to safety recommendations. In part, this is a response to the Chicago Convention Annex 13 requirement [slide 4] from 1 November 2001 to analyse the information contained in accident and incident reports and occurrence databases to determine any preventative actions required. Our aim is that in time, the ATSB's capacity in this area will be as well regarded as our role in national road safety statistical analysis and research [slide 5]. So far we have released aviation safety analysis reports on bird strikes, and on fuel exhaustion and starvation [slide 6] and summaries are in your folders.

As the ATSB builds a business case to replace our current OASIS aviation occurrence database, our aim will be to place more data on the web and to better integrate with CASA, Airservices, Defence, NZ and the broader Australian industry, while maintaining necessary confidentiality. As required by the Government, the ATSB will also review cost recovery arrangements throughout the bureau.

Without detracting from the important roles of others, I would argue that the ATSB, like BASI in the past, makes a significant contribution to aviation safety [slide 7]. In company with Canada's TSB, we are often most satisfied when we can report positive safety action in our final investigation reports instead of needing to make formal recommendations. However, recommendations are still required and the ATSB made 62 aviation safety recommendations last financial year. They are listed on our website (www.atsb.gov.au) together with all responses but let me give a few examples.

In relation to a Boeing 777-300 [slide 8] engine failure in November 2001 involving a Rolls Royce Trent 800 engine, based on metallurgy failure analysis in the ATSB laboratory [slide 9], the ATSB recommended that Rolls Royce revise a service bulletin to highlight necessary inspections for cracking failure between the lever and connecting pin of the variable stator vane lever assemblies and that the revision be reviewed by the UK CAA. These recommendations were accepted and pleasingly the ATSB's February 2003 final report was featured in a recent Flight Safety Foundation Bulletin.

Yesterday, as the result of investigation work on a recent Robinson R22 loss of main rotor blade in flight fatal accident, the ATSB released a recommendation to the FAA and Robinson Helicopters to inspect a sample of main rotor blade root fittings to establish the integrity of the adhesive bond in the spar to root fitting joint. The R22 blade lost in flight showed evidence of adhesive disbonding allowing moisture to initiate corrosion and fatigue cracking in the region of the inboard bolt hole of the blade root fitting. The ATSB is carefully reviewing the R22's operating history including environmental conditions and is seeking similar data from Robinson. The FAA, Robinson, and the NTSB are taking our recommendation very seriously.

Following a runway incursion involving a Boeing 737 and a sweeper vehicle at Perth Airport [slide 10], the ATSB recommended that Airservices Australia, in conjunction with airport owners and CASA, review the adequacy of equipment, standards and procedures for drivers of vehicles using airport runways. Airservices has responded that among other changes, vehicles that enter and remain on a duty runway will be capable of monitoring and communicating on the tower control frequency and will be controlled by air traffic controllers. Perth, Adelaide and Melbourne control towers are expected to implement the change this month with Canberra soon to follow.

As an example of a recent ATSB serious incident investigation report, let me mention a microburst windshear incident involving VH-TJX, another Boeing 737-400 [slide 11]. On approach into Brisbane at about 500 feet, the weather deteriorated rapidly and the pilot in command initiated a go-around. Shortly after, climb performance of 3600ft/min dropped to 300ft/min due to the combined effects of microburst downdraft and heavy rain. Maximum engine thrust was applied and the aircraft successfully diverted to Maroochydore. In this case, the Bureau of Meteorology public forecast of severe thunderstorms was not provided to the flight crew or to Airservices Australia and air traffic control tower concerns about the intensity of the approaching thunderstorm were also not conveyed to the TJX crew until the aircraft was on final approach. The serious incident highlighted that in the absence of extensive Doppler weather radar capabilities and wind shear detection systems, there is a need for better collaborative decision-making among forecasters, controllers, pilots and operators during periods of severe convective weather.

An important example of a recently released report on an accident investigation involved an approach and landing CFIT accident with the death of an RFDS pilot in a Beech 200C at Mt Gambier [slide 12]. The Bureau found that in the dark night conditions at an unfamiliar aerodrome, for reasons that could not be ascertained but possibly related to self-imposed time pressure, the pilot did not comply with published instrument approach procedures and impacted terrain 3.1 nautical miles short of the runway close to the extended centreline. This again highlights the importance of good CFIT/ALAR awareness. The ATSB report noted the good work of the Aviation Safety Foundation of Australia in conducting Flight Safety Foundation ALAR courses.

Of the final aviation investigation reports released last financial year, the one I am most proud of is our report into Ansett 767 maintenance and continuing airworthiness of Class A aircraft released on 15 November last year [slide 13]. I believe that the report is important on a number of levels and I can only skim the surface here. You may know that the ATSB's investigation was triggered by the groundings of Ansett 767 aircraft in December 2000 and April 2001 primarily as a result of missed fatigue crack inspections. While this action by the operator and regulator assured passenger safety, our investigation uncovered deeper systemic issues. (Locals will recall that the possible consequence of unmanaged fatigue cracking was well-demonstrated by the Vickers Viscount [slide 14] loss of right wing in flight accident in which all 26 on board perished on a New Year's eve flight from Perth to Port Hedland in 1968.)

The ATSBs Ansett investigation [slide 15] found that in addition to errors and omissions by individuals associated with the operator, there were deeper system and resource weaknesses in the airline group and shortcomings by the FAA, both of which CASA was unaware. Ansett omitted to action 25,000 flight cycle inspections issued by Boeing in June 1997 and updated in June 2000 to include fatigue crack inspections [slide 16] of the aircraft tail [slide 17]. It also failed to action within the recommended six months a March 2000 Boeing Alert service bulletin relating to possible cracking [slide 18] in 767 engine mount fittings [slide 19]. The FAA did not mandate Boeings June 1997 inspections for older 767s and subsequent service bulletins until after the second Ansett groundings. Boeing did not highlight the potential safety significance of the tail cracking issue in its service bulletin until November 2001 prior to this the focus was on it being a commercial issue.

Australia's former CAA from 1990 to 1991 had reduced the regulator's in-house capacity to review important manufacturer service bulletins with safety consequences and relied on foreign type regulators and operators to do so. CASA was unaware of delays in the FAA and did not appreciate the extent of problems involving Ansett's maintenance [slide 20]. Vulnerability was compounded by weaker than desirable feedback systems for maintenance issues. The ATSB made a number of recommendations to improve information flows in line with Annexes 8 and 6 of the ICAO framework.

Robust regulatory and operator systems for ageing aircraft built on 'damage tolerance' principles are essential. This aspect is also highlighted in a 2003 paper by Wilson and Locket. When major changes occur to the commercial, organisational, IT and regulatory environments, old ways of doing things need to be carefully re-examined.

As Bruce Gemmell noted yesterday, a recent ATSB discussion paper on Australia's aviation safety in the decade to December 2002 reports substantial improvement [slide 21]. For example, both the accident and fatal accident rate per 100,000 hours flown (and per 100,000 departures) have roughly halved, driven mainly by a reduction in general aviation accidents. High capacity regular public transport has experienced no fatalities since the 1960s and low capacity RPT fatality rates remain low with no significant trend. However, a bad year for GA fatal accidents and certainly one major RPT fatal accident would change the picture as Bruce illustrated. There is no room for complacency - only by 'maintaining the focus' on accident prevention and safety culture can this improvement be sustained and bettered. As our Ansett report emphasised, this requires 'organisational mindfulness' and sound communication and cooperation within Australia and internationally including among manufacturers, operators, ATC, regulators, and investigators.

The ATSB continues to exercise its response capability against the possibility that there is a major RPT fatal accident in Australia. An on-site scenario exercise called 'POPFLOT' will be occurring in the near future. Where resources permit, we continue to send investigators to assist with major accidents overseas to gain experience [slide 22]. Most recently we did this with the Ilyushin IL-76 TD fatal accident in East Timor (where I acknowledge gratefully the support by DFS-ADF and DSTO), the SQ006 747-400 accident in Taipei and the American Airlines A300-600 accident in New York. The ATSB has been working with Emergency Management Australia on a new umbrella Commonwealth/State plan for a major aviation disaster called 'AVDISPLAN' which is close to release. We continue to value our memberships of international bodies such as the Flight Safety Foundation ITSA, and ISASI.

Recent books by Adair, and by Cobb and Primo, provide a very useful insight into likely media and political factors that could follow a major fatal aviation accident. Adair's 2002 book traces the NTSB's five year investigation of the 1994 USAir flight 427 737 crash near Pittsburgh [slide 23] and its ultimate identification of a safety deficiency in the rudder power control unit that could lead to a reversal under unusual conditions. Along the way are all the usual problems with media speculation and public conspiracy theories, blaming the airline and the regulator, and the pilot's union blaming the aircraft while the manufacturer blamed the pilots. In addition, we get internal tensions within the NTSB including between investigators and the Board. I commend it to you as a good read.

Cobb and Primo's 2003 book is more academic and appealed to me as an escaped social scientist. They argue that major aviation accidents in a country lead to media, political and policy responses out of proportion to their statistical significance. They say [slide 24]: Plane crashes capture our attention because they bring to the fore a fact about flying that is often unexpressed: once the cabin door closes, passengers are at the mercy of the crew and the equipment. By nature, humans are loath to relinquish control over their fate & As a consequence, faith in the air travel system is contingent on the public learning the reason a plane crashed & 'In such a culture the ultimate horror is a disaster without an explanation, an essentially random event'. Cobb and Primo examine five hypotheses that lead to increased media coverage proxied as the dependent variable by New York Times articles. The five are [slide 25]: increasing number of deaths; proximity to a major metropolitan area; increasing level of causal uncertainty; suspected terrorism or sabotage; and the involvement of 'political entrepreneurs'. My guess is that these factors could be pretty relevant in Australia.

Getting back to the practical side of investigation, I can report that in 2002 the ATSB received tertiary accreditation for five years for a new Diploma in Transport Safety Investigation [slide 26]. This is an in-house competency-based qualification linked to work level standards that indicates that an appropriate standard has been reached by our investigators in a range of areas. On-the-job mentoring and coaching is a key part of the Diploma.

The ATSB is grateful that the Australian Government sponsored new multi-modal safety investigation legislation [slide 27] for aviation, marine and rail which received bipartisan support in the Parliament. The Transport Safety Investigation Act 2003 and accompanying Transport Safety (Consequential Amendments) Act 2003 and Transport Safety Investigation Regulations 2003 came into effect on 1 July. The new TSI Act reinforces the fact that the ATSB conducts no-blame aviation safety investigations in accordance with the provisions of Articles 26 and 37 and Annex 13 to the Chicago Convention. The Government's wish for the ATSB to cooperate with other legitimate investigation bodies is made explicit but where necessary the TSI Act provides the ATSB with primacy in investigation unless there is a clear case of 'unlawful interference', such as in a case of terrorism. The TSI Act [slide 28] also reinforces the ATSB's operational independence subject only to a Ministerial power to require the Bureau to initiate an investigation. The Act includes a new power to apply an evidence protection order and stiff penalties for breaching such an order or hindering an investigation. It also bolsters the protection given to sensitive ATSB investigation material and reports, including against use in criminal or civil proceedings other than coronial inquiries. This is to encourage a free flow of safety information to the ATSB in the interests of future safety. Other bodies that determine blame or liability such as CASA or the police operate in parallel and, unlike the ATSB, cannot compel witnesses to give evidence that may self-incriminate.

The ATSB strongly supports ICAO's GASP [slide 29] and will be involved in the ICAO audit process that for the first time in 2004 includes Annex 13. Australia responded to an initial ICAO questionnaire on 31 August 2003. While we will always investigate international accidents occurring in Australian territory as required by Article 26, and we also seek to attend all non-sport fatal aviation accidents where practicable, Australia has notified differences against investigating all domestic accidents in the intensive manner suggested by Annex 13. We believe that this is often not the best use of scarce safety resources which may be better directed towards investigation of serious incidents or to research and analysis of safety trends.

In Australia's federal system, primary responsibility for determining cause of death lies with state and territory coroners who also typically oversight forensic and pathology services. Coroners have an important and difficult job and face pressure to assist loved-ones of the deceased with closure and to resist attempts by counsel to make a name for themselves in local media or to position their clients for future litigation. The ATSB [slide 30] has sought increased understanding and cooperation with coroners and supported their privileged access to investigation information under the TSI Act. Professional relationships with most coroners are good and there is an awareness that the ATSB investigates with the Annex 13 primary goal of future safety, not to expend scarce funds to assess every possibility that may be of interest in coronial or subsequent legal proceedings. In complex investigations with limited evidence, it is always possible to say with the benefit of hindsight that the ATSB could have done better, particularly with greater resource expenditure. However, two recent inquests have, in my opinion, been excessively harsh in their criticism.

The Western Australian State Coroner's findings last September in relation to VH-SKC, the chartered Beech King Air so-called 'ghost flight' which ultimately crashed near Burketown [slide 31], followed most of the ATSB report (that suggested likely hypobaric hypoxia from unknown causes incapacitating the pilot and passengers) and adopted our suggested safety recommendations, but criticised the Bureau for its Annex 13 processes including confidentiality of investigation information such as witness statements and internal investigator analysis. The Coroner concluded that the ATSB had initial control over all relevant exhibits whereas the bodies of the deceased had been removed by state police and pathologists before the ATSB arrived at the accident site. In addition to his safety recommendations, the ATSB welcomed the Coroner's findings that it ultimately appears that the ATSB report was based on a substantial amount of scientific investigation and in relation to pathology testing that the various Coronial jurisdictions clearly have a role to play in this context to ensure that sensible cooperation can take place. But of course, this was not the media's emphasis.

The South Australian State Coroner's findings on 24 July in relation to the VH-MZK Whyalla Airlines Piper Chieftain crash into Spencer Gulf [slide 32] were rather colourful in their criticism of the ATSB's investigation. As many of you know, after the ATSB completed its report in December 2001 the aircraft engine manufacturer began to issue service bulletins pointing to a possible manufacturing problem with the steel in Chieftain crankshafts and VH-MZK's left crankshaft was ultimately included in a September 2002 bulletin. Despite the circumstantial appeal of this development, when we eventually got the crankshaft back from US civil litigation proceedings and destructively tested it, we confirmed our original finding that the crack in the crankshaft was initiated about 50 flights before the crash as a result of the edge of a bearing rubbing on the surface and initiating a thermal crack just below the nitrided case. We found no evidence of a manufacturing defect in the steel that under normal operating conditions could have initiated the crack. An ATSB supplementary report addressing significant evidence since our 2001 report should be available this November.

Despite our bruising in these recent inquests, I remain hopeful that all coroners will work more cooperatively with the ATSB and vice versa under the new TSI Act umbrella. To this end, we are expecting consolidated coronial comments back next month on a proposed memorandum of understanding or declaration of intent.

The ATSB works in a highly charged environment where accidents can have major consequences for the families of deceased, for reputations and financially. While in systemic investigations we are sometimes critical of operators, regulators and service providers and not just the human at the sharp end, we seek to do so in a manner that promotes learning and future safety [slide 33]. We recognise that in a 'just culture' the relatively few deliberate acts that undermine safety should not go unpunished, but that is not the ATSBs role. While there will always be critics, the ATSB has been supportive of CASA's move towards system-based audits and increasing recognition of the human factor. The Government has legislation in process to create a new confidential aviation self-reporting scheme (ASRS) to be administered by the ATSB in place of the CAIR scheme that has served us well since 1988. Similar to the US ASRS, the proposed Australian ASRS scheme will enable those reporting a safety breach to the ATSB under a threshold of seriousness, to claim indemnity from regulatory action by CASA once every five years.

Let me conclude by saying that if in two years time our esteemed Executive Chairman Peter Lloyd, and Bob Warn, Rob Lee and all the others who volunteer time to make Safeskies happen go 'once more into the breach', and then invite me back, there are a couple of things on which I hope to report. First, and most importantly, I hope to be able to say that we have successfully waged war on 'normalised deviancy' and the downward trend in aviation accident and fatality rates has continued and there have been no new RPT fatals. Second, I hope to be able to report that safety culture, cooperation and relationships in the sometimes personality-charged aviation industry have improved and that we are all working together toward our common goal of aviation safety.

On behalf of the ATSB [slide 34], thank you for the opportunity to speak with you this morning.

References

• Adair, Bill (2002) The Mystery of Flight 427: Inside a Crash Investigation, Washington: Smithsonian Institution Press.
• ATSB (2001a) Pilot and Passenger Incapacitation, Beech Super King Air 200 VH-SKC, Wernadinga Station, Qld, Aviation Safety Report BO/200003771, Canberra: Australian Transport Safety Bureau, March.
• ATSB (2001b) Piper PA31-350 Chieftain VH-MZK Spencer Gulf SA, (Whyalla Airlines), Aviation Safety Report 200002157, Canberra: Australian Transport Safety Bureau, December.
• ATSB (2002a) Boeing 737-400, VH-TJX, Brisbane, Qld, Air Safety Investigation 200100213, Canberra: Australian Transport Safety Bureau, August.
• ATSB (2002b) Sweeper on the Runway at Perth International Airport, Air Safety Investigation 200102695, Canberra: Australian Transport Safety Bureau, September.
• ATSB (2002c) Cessna Aircraft Company Conquest, Air Safety Investigation Occurrence Brief 200200095, Canberra: Australian Transport Safety Bureau, October.
• ATSB (2002d) Investigation into Ansett Australia maintenance safety deficiencies and the control of continuing airworthiness of Class A aircraft, Aviation Safety Investigation BS/20010005, Canberra: Australian Transport Safety Bureau, November.
• ATSB (2003a) Boeing Co B777 Air Safety Occurrence Report 200105494, Canberra: Australian Transport Safety Bureau, February.
• ATSB (2003b) Raytheon Beech 200C, VH-FMN, (Mt Gambier SA) Air Safety Occurrence Report BO/2001105769, Canberra: Australian Transport Safety Bureau, February.
• ATSB (2003c) Aviation Safety Indicators 2002: A report on safety indicators relating to Australian aviation, discussion paper, Canberra: Australian Transport Safety Bureau, August.
• ATSB (2003d) Annual Review 2003, Canberra: Australian Transport Safety Bureau, (forthcoming October).
• Barnett, A. & Higgins, M.K. (1989) Airline Safety: The Last Decade, Management Science, Vol.35, No.1, January.
• Chicago Convention (1944), Convention on International Civil Aviation, ICAO (as amended).
• Cobb, R.W. & Primo, D.M. (2003) The Plane Truth: Airline Crashes, the Media, and Transportation Policy, Washington DC, Brookings Institution Press.
• Gregoriades, A., Sutcliffe, A. & Shin J-E. (2003) Assessing the Reliability of Socio-Technical Systems, Systems Engineering, Vol.6, No.3.
• ICAO (2001) Annex 13 to the Chicago Convention, 9th edition, ICAO, July.
• TSI Act (2003) Transport Safety Investigation Act 2003; Transport Safety (Consequential Amendments) Act 2003; Transport Safety Investigation Regulations 2003, Commonwealth of Australia: Commonwealth Government Printer, all available on the ATSB website: www.atsb.gov.au.
• Wilson, E.S. & Locket, R.D. (2003) Managing Ageing Aircraft Structures, paper presented at the 10th Australian International Aerospace Congress, Brisbane, 29 July-1 August.