216
Sensemaking in Safety Critical and Complex Situations
( e.g., display, joystick, keyboard, mouse) ( Rogers, Palmer, Chitwood, &
Hover, 2004; Williams, 2004).
• Thompson applied ( N on-DoD) HFACS model to UAS mishap reports. He
found that 60% of all mishaps studied were due to human factors; of these
25% were ascribed to the HF/ E in UAS GCS. His recommendations include
optimization of UAS GCS interface by refining the layout of IO devices to
provide an ergonomic fit for operators ( e.g., display, keyboard, mouse, joystick) ( Thompson & Tvaryanas, 2005).
• Thompson applied the DoD HFACS model to study UAS mishaps. He
found that 50% of mishaps were related to human factors; of those 19%
were ascribed to HF/ E in UAS GCS. His findings included improper clearance and alignment of IO controls ( e.g., display, keyboard, joystick, mouse)
to be the leading cause in HF/ E -related UAS mishaps ( Thompson &
Tvaryanas, 2008).
uaS miShaP inveStigation moDelS for hf/ e evaluation
If the human developers/ operators of t echnology-intensive systems see deficient
or unsafe process or a system, they would not allow it to continue ( Dekker, 2001).
However, they do not foresee the accidents, because they cannot predict the possibility of accident actually happening ( Wagenaar & Groeneweg, 1987). As mentioned
by Dekker ( 2004) human error can be seen as a consequence of deeper issues ( i.e.,
poor design, poor training, mental overload, fatigue) with the system ( Dekker, 2004).
In safety-critical GCS, accidents typically have more than one reason, as they are
a collection of knottily interrelated chain of events. Where these knots need to be
untied delicately to see the actual flow of events leading to an accident, an accurate
and comprehensive post-accident documentation is the most important step that can
result in improving the resilience and safety of the system and possibly halt similar recurrences ( Gordon, Jeffries, & Flin, 2002), preferably with the involvement of
human factors experts.
Therefore, taxonomies/ models are continually being developed and improved by
researchers to methodically collect and document p ost-mishap data for a human factors analysis. Since, varying perspectives on human factors are applied during the
development of such models, it is nearly impossible to select an appropriate model
for capturing HF/ E -related mishap data, because not all models can properly evaluate HF/ E issues ( Andersen et al., 2002; Wiegmann & Shappell, 2001). On the other
hand, some models may be good at capturing IO-related HF/ E data during a mishap
investigation, but may not be able to properly evaluate collected data to find the
underlying HF/ E issues, leading to a missed opportunity to find and eliminate HF/ E
issues retrospectively. A report by the European Organization for the Safety of Air
Navigation provided an overview of models used in aviation mishap investigations
( Andersen et al., 2002):
• Task-based taxonomies: Captures data from the operator’s perspective. IO
HF/ E issues can only be found if an operator realized and highlighted such
shortfalls.
Sensemaking in Safety Critical and Complex Situations
( e.g., display, joystick, keyboard, mouse) ( Rogers, Palmer, Chitwood, &
Hover, 2004; Williams, 2004).
• Thompson applied ( N on-DoD) HFACS model to UAS mishap reports. He
found that 60% of all mishaps studied were due to human factors; of these
25% were ascribed to the HF/ E in UAS GCS. His recommendations include
optimization of UAS GCS interface by refining the layout of IO devices to
provide an ergonomic fit for operators ( e.g., display, keyboard, mouse, joystick) ( Thompson & Tvaryanas, 2005).
• Thompson applied the DoD HFACS model to study UAS mishaps. He
found that 50% of mishaps were related to human factors; of those 19%
were ascribed to HF/ E in UAS GCS. His findings included improper clearance and alignment of IO controls ( e.g., display, keyboard, joystick, mouse)
to be the leading cause in HF/ E -related UAS mishaps ( Thompson &
Tvaryanas, 2008).
uaS miShaP inveStigation moDelS for hf/ e evaluation
If the human developers/ operators of t echnology-intensive systems see deficient
or unsafe process or a system, they would not allow it to continue ( Dekker, 2001).
However, they do not foresee the accidents, because they cannot predict the possibility of accident actually happening ( Wagenaar & Groeneweg, 1987). As mentioned
by Dekker ( 2004) human error can be seen as a consequence of deeper issues ( i.e.,
poor design, poor training, mental overload, fatigue) with the system ( Dekker, 2004).
In safety-critical GCS, accidents typically have more than one reason, as they are
a collection of knottily interrelated chain of events. Where these knots need to be
untied delicately to see the actual flow of events leading to an accident, an accurate
and comprehensive post-accident documentation is the most important step that can
result in improving the resilience and safety of the system and possibly halt similar recurrences ( Gordon, Jeffries, & Flin, 2002), preferably with the involvement of
human factors experts.
Therefore, taxonomies/ models are continually being developed and improved by
researchers to methodically collect and document p ost-mishap data for a human factors analysis. Since, varying perspectives on human factors are applied during the
development of such models, it is nearly impossible to select an appropriate model
for capturing HF/ E -related mishap data, because not all models can properly evaluate HF/ E issues ( Andersen et al., 2002; Wiegmann & Shappell, 2001). On the other
hand, some models may be good at capturing IO-related HF/ E data during a mishap
investigation, but may not be able to properly evaluate collected data to find the
underlying HF/ E issues, leading to a missed opportunity to find and eliminate HF/ E
issues retrospectively. A report by the European Organization for the Safety of Air
Navigation provided an overview of models used in aviation mishap investigations
( Andersen et al., 2002):
• Task-based taxonomies: Captures data from the operator’s perspective. IO
HF/ E issues can only be found if an operator realized and highlighted such
shortfalls.