1. Normal Operations Safety Survey (NOSS)
Chris Henry
The University of Texas Human Factors Research Project
The University of Texas at Austin
2nd ICAO TEM & NOSS Symposium
Washington DC
February 7, 2007
UT Human Factors Research Project © Copyright

2. Presentation Objectives
Why conduct NOSS?
Factors that lead to a successful NOSS
Frequently asked questions
UT Human Factors Research Project © Copyright

3. Aviation Safety Envelope
Accidents
Incidents
Normal Operations
Safety
UT Human Factors Research Project © Copyright

4. Safety Data Coverage Accidents Incidents Normal Operations
Voluntary Incident Reports
NOSS
On any given day airlines may collect information from
This diagram represents sources of aviation safety data
We collect 4 main types of informaiton with the safety envelope:
Accident investigations – rarest, non-voluntary, reactive investigations
FOQA – technical flight parameters
ASAP – self-report data resulting from voluntary crew reporting, proactive, center on FAR violations
LOSA – normal line operations, info from SME’s on normal flights
At UT – we have developed a multi-project research program that covers assessment and analysis of 3 of these data collection programs
- reaching each level of the safety envelope
Accident Investigation
Mandatory Incident Reports
UT Human Factors Research Project © Copyright

5. Aviation Safety Envelope
Accidents
Incidents
Normal Operations
NOSS Rationale
Proactive snapshot of system / controller performance strengths and weaknesses
in normal operations (just like a health checkup)
On any given day airlines may collect information from
This diagram represents sources of aviation safety data
We collect 4 main types of informaiton with the safety envelope:
Accident investigations – rarest, non-voluntary, reactive investigations
FOQA – technical flight parameters
ASAP – self-report data resulting from voluntary crew reporting, proactive, center on FAR violations
LOSA – normal line operations, info from SME’s on normal flights
At UT – we have developed a multi-project research program that covers assessment and analysis of 3 of these data collection programs
- reaching each level of the safety envelope
UT Human Factors Research Project © Copyright

6. What can NOSS provide?
Identify threats within the operating environment
Check the quality/usability of procedures
Identify problems with controller/equipment interfaces
Assess the degree of transference of training to the line
Understand controller shortcuts and workarounds
Make comparisons across units or facilities
Assess safety margins
UT Human Factors Research Project © Copyright

7. NOSS Success Factors
NOSS success is dependent upon methodology and execution
Formal Check
Regulator
Nobody
Angel
Performance
NOSS Observer
Natural
Performance
- Controller Trust + - +
NOSS value - +
Low controller trust = Low quality data because there will be no differentiation between NOSS and proficiency checks
UT Human Factors Research Project © Copyright

8. NOSS Operating Characteristics
Over-the-shoulder observations during normal shifts
No observations of controllers who are undergoing training
No observations while checks are being conducted
Voluntary controller participation
Controllers have a right to decline a NOSS observation
Denial rates in NOSS trials have been low
Joint management / association sponsorship
Letter signed by management & association representatives sent to all controllers
UT Human Factors Research Project © Copyright

9. NOSS Operating Characteristics
Trusted and trained observers
Trust and Credibility
Selection process – management / association list of candidates
Diversity of observers increases buy-in
Training
Classroom training, test observations, recalibration session
Curriculum
TEM principles and exercises
NOSS observation protocols
Narrative guidelines
UT Human Factors Research Project © Copyright

10. NOSS Operating Characteristics
Anonymous, confidential, and non-punitive data collection
No names, employee numbers, dates, experience, or other identifying information
Data used for safety purposes only – no punitive actions
Observers identity is known only by the third party facilitator
Trusted and secure data collection site
Third party or controller association gatekeeper
Controllers and observers must be comfortable with the data collection site
Systematic data collection instrument
No judgments – observers record data based on TEM events
UT Human Factors Research Project © Copyright

11. NOSS Safety Metrics What do controllers do to manage their airspace?
Manage operational complexity 1
Threat Management
Manage their own errors 2
Error Management 3
Manage deviations in airspace status
Undesired State Management
The Threat and Error Management Model serves as the “NOSS Safety Metrics”
UT Human Factors Research Project © Copyright

12. NOSS Operating Characteristics
Data verification process
TEM data checked to ensure coding accuracy and consistency with SOP
Data analysis does not begin until verification has been completed
Data-derived targets for safety enhancement
Serve as benchmarks for safety change
Initial NOSS, safety change process, follow-up NOSS
Feedback results to controllers
Results summarized for controllers
Information on how organization intends to respond to the data
UT Human Factors Research Project © Copyright

13. NOSS: Gaining Controller Trust
Over-the-shoulder observations during normal shifts
Joint management / association sponsorship
Voluntary Participation
Trusted and trained observers
Anonymous, confidential, and non-punitive data collection
Trusted and secure data collection site
Systematic data collection instrument
Data verification process
Data-derived targets for safety enhancement
Feedback results to controllers
UT Human Factors Research Project © Copyright

14. NOSS Defined
The 10 characteristics that differentiate NOSS (LOSA) from other methodologies have been endorsed by the:
Federal Aviation Administration
International Civil Aviation Organization
International Air Transport Association
International Federation of Airline Pilots Association
International Federation of Air Traffic Controllers (6)
The LOSA Collaborative
US Airline Pilots Association
University of Texas
Programs that omit one or more of the characteristics might be useful and beneficial, but they are not NOSS
NOSS must have all ten characteristics
UT Human Factors Research Project © Copyright

15. Frequent Questions How long are NOSS observations?
Observations approximately one hour in duration have given us good data
Do we observe individual controllers/positions or teams?
Team dynamics captured, but from the perspective of an individual operating position
Should NOSS be continuous or a snapshot?
Snapshot allows organization to assess and act on data
Other data sources can inform between snapshots
Little benefit for the increased cost of doing continuous NOSS
What should the scope of a NOSS be?
Depends on the size of the facilities and the comparisons to be made
UT Human Factors Research Project © Copyright