ATSB fatality investigations in
Australia and the Asia Pacific
Kym Bills
Executive Director ATSB
Asia-Pacific Coroners’ Conference
31 October 2007
Overview
•
•
•
•
I am grateful for the opportunity to be here
I will give a short overview of the ATSB
Provide some context on transport fatalities
Indicate how and why the ATSB seeks to
prioritise limited investigation resources
• Outline our recent Asia Pacific involvement
• Introduce the methodology used in larger
investigations such as Lockhart River, and
• Note issues with legal concepts and the
desirability of further Coroners’ dialogue.
Multi-modal
ATSB
ATSB
• Multi-modal body to investigate, analyse and
report independently on aviation, interstate
rail & major marine safety occurrences
• Transport safety investigations are not
intended to be the means to apportion
blame or liability, in accordance with the
Transport Safety Investigation Act 2003 (TSI
Act) & Annex 13 to the Chicago Convention
• Powers to investigate, including to search
and compel evidence even if incriminatory
• But reports/evidence can’t be used in courts
with the exception of Coronial inquests.
• The ATSB is part of the Australian
Government Department of Transport and
Regional Services (DOTARS) for
administrative and resourcing purposes
• Separate from State bodies like Police and
rail regulators, and federal bodies like the
Civil Aviation Safety Authority (CASA) & the
Australian Rail Track Corporation (ARTC)
• Importantly, separate investigations by
police, regulators and OHS bodies occur
consistent with a ‘just culture’ (perhaps 10%
of accidents via a form of culpable actions).
• Hence, the ATSB’s no-blame safety
investigation is only one part of the system
• To reinforce independence, under the TSI
Act S15, neither the Minister nor Secretary
are to influence ATSB investigations
• ATSB mandatory occurrence reporting,
voluntary confidential reporting and data
analysis and research supplement both
investigation & industry schemes, eg SMS
• Approx 115 ATSB staff, most in Canberra
and three quarters aviation-related
including notifications, administration etc
- annual budget now almost $20 million.
Australian transport safety data
• The overall transport accident death rate
across road, rail, marine and aviation
decreased from 10.4 deaths per 100,000
population in 1997 to 8.3 in 2006:
11
Deaths per 100 000 people
10
9
8
7
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
Australian transport safety data
• As in other OECD countries, most fatalities
are on roads– overall crash cost A$18b pa
Year
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
Road
1767
1755
1764
1817
1737
1715
1621
1583
1627
1599
Rail
68
59
47
46
56
59
48
47
38
40
Marine
46
46
51
42
59
48
41
41
39
na
Aviation
37
55
46
43
42
34
44
33
43
41
Road safety fatality data
• Major factors remain speed, alcohol/drugs,
fatigue, distraction, vehicle design and
safety features, road design and features
• Most deaths occur among car drivers and
passengers and disproportionately among
young post-P plate drivers
• Motorcyclist fatalities up 22% since 2004
and growing among over-40 males but most
deaths still young males
• About 30,000 serious injuries on our roads
annually (via hospitalisation), c2/3 males.
Road safety data
• ATC National Road Safety Action Plan for
2007 & 2008 coordinated by the ATSB
includes many cost-effective measures:
- that have not been implemented, or
- have not been implemented in all
jurisdictions, or
- have not been implemented on a
sufficient scale
• eg Victoria’s tough speed campaign:
Victorian campaign on speeding:
Rolling 12 month average
11
Death rate per 100 000 population
10
9
Australia
(excluding Victoria)
8
7
Victorian road deaths fell
by 31% in the two years
to April 2004
Victoria
6
5
Dec-98
Dec-00
Dec-02
Dec-04
Dec-06
Dec-08
Marine safety data
• Mostly via recreational boating there are
about 40 marine fatalities annually
• The ATSB mainly investigates large ships
(also Cwth vessels like Malu Sara) and the
almost 250 investigations since 1991
include 34 fatal accidents, 64 groundings,
41 collisions, and 28 fires/explosions
• A number of fatalities involved lifeboat drills
and collisions with fishing vessels
• ATSB key role at the IMO in improving
investigation requirements/standards.
Rail safety data
• ATSB has been working with state and NT
rail regulators since 1999 to obtain and
publish nationally comparable safety data
• After many false dawns, in June 2007 we
published regulator data covering calendar
2001 to 2006 inclusive
• This is a basis to build from via National
Transport Commission (NTC) and the
Australasian Railway Association (ARA).
Rail safety data
• National regulator data for calendar 2006
include:
- 40 ‘non-suicide’ deaths
- 127 serious injuries (excluding NSW)
- 120 running line derailments
- 191 running line collisions (105 with
infrastructure, 46 with persons, 16 road
vehicles, 14 rollingstock, 10 other trains)
- 92 level crossing collisions (82 with road
vehicles, 10 with persons)
- 414 serious ‘signals passed at danger’.
Rail safety investigations
• ATSB investigation reports have led to
significant safety action in the areas of
medical standards, vigilance devices/pilot
valves, the need for better/standardised
communications, human error-tolerant
systems, and level crossing risk-based
treatments and broader education
• There remain a number of serious level
crossing accident investigations underway
Aviation safety data
• Fatal LCRPT accidents in 2000 (Whyalla)
& 2005 (Lockhart River) but most
accidents/fatalities GA (non-sport) shown:
250
Fatal Accidents
Non-fatal Accidents
200
All Accidents
150
100
50
0
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
Aviation safety data
• We continue to see many of the same
types of fatal accident, eg:
- controlled flight into terrain
- weather (eg VFR into IFR conditions)
- fuel exhaustion/starvation
- hitting powerlines
- high risk GA behaviour (eg low passes)
• In aviation & other modes human factors
continue to dominate - management lack
of awareness of human performance limits
and structures remains an issue.
Some key ATSB references
• ATSB website has: an interactive road
safety database and monthly analysis;
initial rail safety data for 2001-2006;
aviation data updated monthly; basic
marine data in the ATSB Annual Review
• ATSB research reports for road & aviation:
eg Digest of all ATSB research for 2006;
HFACS comparison with US; initial
Australian Aviation Safety in Review
• (n.b. we use Coronial database especially
in road and road/rail safety research.)
Investigation prioritisation
• In aviation we received over 13,000 event
reports last FY of which almost 8,000 were
classified as accidents (112) or incidents
• Resourced for 80 investigations (30 larger)
• Required under Chicago Convention to
investigate international carrier accidents
• Annex 13 standard re all accidents and
recommended practice re serious incidents
• Do non-sport fatals to assist Coroners but
often most safety value in selected RPT
incidents so very tough judgements/choices
Asia Pacific role
• We regularly assist the NZ TAIC and the
ADF especially with aviation investigations
• Resources are tight and eg PNG has no
legislation and there are sensitivities re Fiji
• PASO regulation may lead to investigation
• Helped Taiwan ASC set up recorder lab &
do read-outs when requested
• Teach in region - Singapore, Hong Kong
• Investigations include QF1, SQ006, IL76:
QF1 747-400
Bangkok
SQ006 747-400, Taipei
IL76 CFIT
Bacau, East
Timor
IL76 CFIT Bacau,
East Timor
Indonesia safety cooperation
• The ATSB has a longstanding relationship
with Indonesian counterparts (NTSC) and
had arranged last year a pilot placement
• Garuda 7 March 2007 accident we
assisted on site and with recorders
- major report drafting assistance et al
• Ongoing cooperation re Adam Air 1/1/07
and marine investigations, plus training in
investigation, HF etc
Garuda 737-400, Yogyakarta
Based on
professionalism
and sound
methodology
this is what we
seek to avoid!
Safety Investigation
200501977
Collision with Terrain
11 km NW Lockhart River Aerodrome
7 May 2005, RPT 2 crew/13 pax fatalities
VH-TFU, SA227-DC (Metro)
Lockhart River Investigation
Large team of investigators:
• Examined aircraft components - on-site and
in laboratory
• Examined cockpit voice recorder and flight
data recorder and radio recordings
• Hindered by lack of usable cockpit voice
recorder information, level of damage to
aircraft, and no survivors or witnesses
• Assisted by flight data recorder, 25,000
pages of evidence, over 100 interviews
Lockhart River investigation
• Considered all aspects of the aviation
system which included organisational &
regulatory issues as well as aircraft/crew
• Conducted a research study into
instrument approaches
• 500 page final report released on 4 April
2007 used ATSB Safety Investigation
Information Management System (SIIMS)
• May 2004 $6.1m Australian Government
Budget funding for SIIMS over 4 years.
Controlled flight into terrain (CFIT)
• Pilots lost situational awareness in cloud
during a type of GPS-linked approach &
made no discernable attempt to pull up
• No evidence of
– problems with flight controls
– problems with engines or propellers
– pilot incapacitation
– windshear etc.
Lockhart River approach profile
South Pap
Accident site
ATSB investigation analysis model
Risk Controls
(Recovery)
Incident
Individual
Actions
Organisational
Influences
Risk Controls
(Preventive)
Local
Conditions
Production
Goals
Technical
Events
Risk Controls
(Recovery)
Accident
ATSB investigation analysis model
Contributing safety factors
• Defined as a safety factor that, if it hadn’t
occurred/existed … the accident would
probably not have occurred … or another
contributing safety factor would probably
not have occurred or existed
• Evidence not sufficient for some (eg CRM)
with ‘probably’ defined as >66% (c. 2 in 3)
• ‘Acci-map’ diagram shows 19 contributing
safety factors (black border) and 13 other
safety factors (purple outline).
Regulatory
Oversight
Organisational
Influences
CASA guidelines
for inspectors
Transair
organisational
structure
CASA AOC
approval
processes
Consistency with
CASA oversight
requirements
Transair chief
pilot commitment
to safety
CASA airline
risk profiles
Pilot checking
Crew
endorsements
and clearance
to line
Local
Conditions
Common
practices of pilot
in command
Supervision of
flight operations
Pilot training
Copilot ability for
the RNAV (GNSS)
approaches
Operations
manual SOPs
for approaches
Conducting RNAV
(GNSS) approach when
copilot not endorsed
CRM conditions
Cockpit layout
Collision with Terrain
11 km NW Lockhart River Aerodrome
7 May 2005
VH-TFU, SA227-DC
Descent speeds, approach
speeds and rate of descent
exceeded
RNAV
approach
waypoint
names
Approach chart
design issues
High workload
Individual
Actions
CASA processes
for accepting
approaches
Transair risk
management
processes
Operations
manual
useability
Risk
Controls
CASA processes
for evaluating
operations manual
Regulatory
requirements
TAWS not
fitted
GPWS on
normal
approaches
Runway 12
RNAV (GNSS)
approach design
Loss of
situational
awareness
Descent below
segment minimum
safe altitude
Descent
problems not
corrected
Controlled flight
into terrain
The Acci-map diagram is built from bottom up
Individual
Actions
Conducting RNAV
(GNSS) approach when
copilot not endorsed
Collision with Terrain
11 km NW Lockhart River Aerodrome
7 May 2005
VH-TFU, SA227-DC
Descent speeds, approach
speeds and rate of descent
exceeded
Descent below
segment minimum
safe altitude
Descent
problems not
corrected
Controlled flight
into terrain
Regulatory
Oversight
Organisational
Influences
CASA guidelines
for inspectors
Transair
organisational
structure
CASA AOC
approval
processes
Consistency with
CASA oversight
requirements
Transair chief
pilot commitment
to safety
CASA airline
risk profiles
Pilot checking
Crew
endorsements
and clearance
to line
Local
Conditions
Common
practices of pilot
in command
Supervision of
flight operations
Pilot training
Copilot ability for
the RNAV (GNSS)
approaches
Operations
manual SOPs
for approaches
Conducting RNAV
(GNSS) approach when
copilot not endorsed
CRM conditions
Cockpit layout
Collision with Terrain
11 km NW Lockhart River Aerodrome
7 May 2005
VH-TFU, SA227-DC
Descent speeds, approach
speeds and rate of descent
exceeded
RNAV
approach
waypoint
names
Approach chart
design issues
High workload
Individual
Actions
CASA processes
for accepting
approaches
Transair risk
management
processes
Operations
manual
useability
Risk
Controls
CASA processes
for evaluating
operations manual
Regulatory
requirements
TAWS not
fitted
GPWS on
normal
approaches
Runway 12
RNAV (GNSS)
approach design
Loss of
situational
awareness
Descent below
segment minimum
safe altitude
Descent
problems not
corrected
Controlled flight
into terrain
Lockhart River methodology
• QLD State Coroner used the ATSB Final
report and research report as key evidence
• CASA suggested methodology not mature,
biased towards finding regulator issues,
>66% test too weak etc
• No direct influence on cockpit
• Coroner raised other methodology issues
re 66% being too precise or mechanistic &
the desirability of a ‘Briginshaw’ type test, ie
tougher on evidence if grave consequences
Methodology paper
• ATSB had been criticised in the Whyalla
inquest that not precise enough re
probability terminology, and methodology
• Double engine failure sequence: did use
judgment as well as statistical probabilities
• While judgement also used in assessing
the >66% (cf US ‘clear & convincing’) and
probably/likely terminology is supported by
International Panel on Climate Change, I
have commissioned a paper on the issues
Methodology paper
• Perhaps Coroners’ end-focus on fatalities
(>50% causality) & safety issues additional cf
ATSB focus at each link (>66%) via broad
view of causality & safety issues regardless
of contribution may occasion differences
• Issues in legal writing & case law regarding
use of probabilities and Briginshaw
• McHugh J (1995) “I know Briginshaw is cited like
it was some ritual incantation. It has never
impressed me too much. I mean, it really means no
more than, ‘Oh, we had better look at this a bit more
closely than we might otherwise’, but it is still a
balance of probabilities in the end.”
Methodology paper
• Anderson et al (CUP 2005) “Until recently
most legal scholars have accepted the view that
evaluation of evidence can rarely be governed by
rules. … The so-called ‘probability debates’ in law
have been underway now for over thirty years.”
• Freckleton et al note pressures at inquests
by parties seeking to prepare for other
litigation can make them adversarial etc
• Hopkins: focus varies based on self-interest
• My aim is that a draft paper initially be sent
to State Coroners for comments
Growing challenges
• Striking the right balance between
protecting safety data and regulatory and
legal systems
• Getting the balance right between noblame and culpability in a ‘just culture’
• Professionalism and robust methodology
• What trade-off should there be between
investigation timeliness and thoroughness
(eg with media and societal expectations)
• The growing safety/security interface
Continuing challenges
• Using all available means to avoid a major
accident is a primary challenge
• This includes good safety management
systems among all key players
• Understanding of the limits to human
performance and organisational behaviour
• Risk analysis, threat & error management
• Helping foster safety cooperation and
excellence in regulation
• Learning from others, mindfulness of past
lessons, & never taking safety for granted:
SAFETY
MANAGEMENT
AND THE
SINKING OF
BRAZIL’S
LARGEST
OFFSHORE
OIL
PLATFORM
March 2001
COURTESY OF: PAT
HUMISTON & GEOFF DELL
A lesson against hubris &
for those involved in
cutting project
management costs
The following quote is from one of
the company’s financial executives
extolling the benefits of cutting
safety management, quality
assurance & inspection costs
on the P36 project that
sunk into the Atlantic
Ocean off the coast of
Brazil in March 2001
“The Company has established new global
benchmarks for the generation of exceptional
shareholder wealth”
“through an aggressive and innovative
program of cost cutting
on its P36 production facility.”
“Conventional constraints
have been successfully
challenged”
“and replaced with new paradigms appropriate
to the globalised corporate market place.”
“Through an integrated
network of facilitated
workshops”
“the project successfully rejected the established
constricting and negative influences of prescriptive
engineering onerous quality requirements, and outdated
concepts of inspection and client control ”
“Elimination of these unnecessary straightjackets has
empowered the project's suppliers and contractors to
propose highly economical solutions”
“with the win-win bonus of
enhanced profitability margins
for themselves.”
“The P36 platform shows the shape of things to come...”
“in the unregulated global market economy
of the 21st Century”
Humility, mindfulness and common sense
are required for future safety & willingness
to learn from history & each other
Thank you
Questions?