Kristiansen Svein. Maritime Transportation: Safety Management and Risk Analysis
Подождите немного. Документ загружается.
management. It even influences our routine analysis of accidents. We are still waiting
to see professional accident investigation and analysis on a broad scale among even the
advanced shippin g nations (Caridis et al., 1999).
11. 3 ACCIDENT CHAR ACTERISTICS
In the effort to increase safety an d in particular to minimize the effect of human error,
we should keep the following lessons about accidents in mind:
1. Routine:
. Often related to normal operational situations
. Not necessarily a result of abnormal events or conditions
2. Gradual escalation:
. Seldom happens momentarily
. Inability to cope with events as they surface
3. Multiple causes:
. A combination of more than a single event
. Causal factors are related to technology, humans and organization
4. Human error:
. The operator is by definition near the accident events
. The human behaviour should be viewed in broader scope
5. Presence of situation factors:
. Physical environment
. Workplace conditions
. Task load
. Mental and motivational state
It is important to have a realistic view of the time scale for the events related to an
accident, as illustrated in Table 11.1. Decisions that may lead to hazardous conditions
can be taken in the order of a year in advance. The critical events may develop within
a day or an hour and the dramatic release of energy in the order of seconds. In a
corresponding manner, the consequences in terms of breakdown of vessel, human
suffering and environmental damage follow in a dramatic way but also have long-
range effects. This perspective is important to keep in mind when we attack the human
factors problem.
11.4 HUMAN INFORMATION PROCES SING
11.4.1 Accident-ProneTasks
One may ask oneself whether specific tasks are more subject to error than others. This
was investigated in the US nuclear reactor study (RSS, 1975). One of the lessons to be
316 CHAPTER 11 H UMAN FACT O RS
learned was that non-routine tasks such as testing/calibration and repair/modification
were especially prone to omission error, as can be seen from Table 11.2. The analysis
of behavioural mechanisms showed that functionally isolated acts often lead to error
(Table 11.3). This mechanism is typical for non-routine tasks.
Rasmussen (1982) makes an important distinction between the following basic
performance levels: skill-, rule- and knowledge-based behaviour (see Figure 11.2). Skill-
based behaviour is our learned and almost unconscious actions. For certain familiar
situations we apply rule-based behaviour. The complex situations require some kind of
problem solving called knowledge-based behaviour. Skill-based behaviour is most
frequent in daily operations and also less subject to error and accidents. On the other
hand we apply knowledge-based behaviour to deal with unfamiliar and difficul t tasks.
This behaviour may also have the largest consequences in case of error.
Ta b l e 11 . 2 . Omissions in nuclear power plant operations by task category
Task Omissions (%)
Monitoring and inspection 0
Supervisory control 2.3
Manual operation and control 5.9
Inventory control 9.4
Test and calibration 32.9
Repair and modification 41.2
Administrative task 1.2
Management, staff planning 1.2
Other or not specified 5.9
All tasks 100
Source: RSS (1975).
T able 11.1. Characte rist ic ti me scale of a sh i p accident
Time scale Typical event
1 year Management decisions related to operations
1 day Breakdown of safety function
1 hour Initiating failure or error
1 min Attempt to avoid threatening accident
1 sec Release of energy
1 min Collapse of hull and breakdown of systems
1 hour Flooding and fire, loss of vessel and evacuation of people
1 day Rescue of people and vessel
1 week Pollution, hospitalization of people
1 year Environmental damage, post-traumatic effects on people
11 .4 HUMAN INFORMATION PROCESSING 317
11 .4.2 A H uman Control Model
Operation of vessels is still to a large degree based on human–machine control. Figure 11.3
presents a conceptual model of vessel control that outlines key topics relating to human
factors that will be discussed in the rest of this chapter. The core of the control loop is the
interaction between the operator function and the vessel.
The operator uses data from the visual en vironment and the information displays
onboard. This data is processed and results in control actions and communication with
Ta b l e 11 . 3 . Errors in nuclear power plant operations by behavioural mechanism
Behavioural mechanism Errors (%)
Absent-mindedness 1.5
Familiar association 3.0
Alertness low 5.0
Omission of functionally isolated act 34.0
Other omissions 8.5
Mistake among alternatives 5.5
Strong expectation 5.0
Side effect not considered 7.5
Latent condition not considered 10.0
Manual variability, lack of precision 5.0
Weak spatial orientation 5.0
Other not specified 10.0
All behaviours 100
Source: RSS (1975).
Figure 11.2. Alternative human performance levels. (Adapted from Rasmussen,1982.)
318 CHAPTER 11 H UMAN FACT O RS
other crew members. The control of the vessel may, however, be disturbed by the physical
environment. It should also be kept in mind that the performance of the operator is
governed by individual characteristics and the interaction with supervisors and other crew
members.
11. 5 SENSING AND OBSERVATION
Navigation tasks are quite demanding for the officer on watch and the lookout. The key
sources of information are:
. The seaway, landscape, sea marks and marine traffic (marine environment)
. The visual and aural displays on the bridge
. The behaviour of the vessel (movement, acceleration, etc.)
11. 5 .1 Ni g ht Vis io n
Although all vessels are equipped with radar, visual lookout is still an important method
for assessing the position of the vessel in restricted seaways and detecting and monitoring
the traffic. The ability to observe under night (dark) conditions is especially critical.
Figure 11.3. Human ^ machine control of a vessel.
11.5 SENSING AND OBSER VA TION 319
It is well known that many collision accidents happen just after the watch has been relieved
due to inadequate dark adaptation. Madison (1974) states that it takes in the order of
25 minutes to fully adapt to darkness following exposure to sunlight or artificial light.
This can be explai ned by the fact that the eye has two types of photoreceptors, namely
rods and cones:
. Cones react to higher light levels, observation of finer details and perception
of colours.
. Rods are associated with night vision and low-intensity light. They are relatively
insensitive to red light, which is in the lowest frequency of the visual spectrum.
Regardless of frequency of stimulus, objects are seen in shades of black and white.
Figure 11.4 shows that the dark adaption curve is discontinuous. The cones adapt in
10 minutes whereas the sensitivity of the rods is rapidly lost in high-level illumination and
therefore need in the order of 25 minutes to adapt.
11.5.2 TheUniqueRoleofRedLight
Under low-level illumination at night we are dependent on the rods. As they are relatively
insensitive to red light, one may speed up the dark adaptation by wearing red goggles.
The rods are not stimulated by low-level red light and start to adapt as if they were in
total darkness.
In order to minimize the effect on night vision instruments on the bridge may use red
illumination. This solution, however, has certain drawbacks. First, visual acuity is poorer,
which means that the ability to discriminate finer details is reduced. Secondly, visual
fatigue will increase. In certain instances where visual acuity is critical blue lighting is
applied, as for instance on radars and CRT screens. This on the other hand will reduce the
night adaptation.
Both the use of red and blue lighting may also affect the reading of charts and this
should therefore also be taken into consideration. Red, orange and buff colours have a
tendency to disappear under red lighting.
Figure 11.4. Dark adaptat ion curve ( Jay le et al.,1959).
320 CHAPTER 11 HU MAN FA CT ORS
11.5.3 Lookout
Lookout is often a challenging task for various reasons. One factor is that at open sea
there are few objects to observe and this may lead to reduced vigilance. There have been
extensive discussions of what is the best lookout strategy (Madison, 1974):
. Slow rowing gaze, or
. Repeated fixations at widely separated positions
Systematic search seems to be most efficient for low contrast targets whereas free search is
best for high visibility targets. It is also a well-known fact that one should not stare directly
at a dim spot under dark conditions in order to avoid the interference of the blind spot in
the eye. Night light sensitivity is in fact greatest 15–20
away from the centre of vision.
The lookout function may also be degraded by night blindness. The existence or degree
of night blindness may be unknown to the person itself as it is not apparent during
daytime conditions. The source may be either psychological or physiological. It has been
related to neuroses, psychoses and hysteria (Jayle et al., 1959). Other pathological factors
that result in night blindness are dietary deficiencies (vitamin A), diabetes, glaucoma or
congenital night blindn ess. It has also been shown that ageing is an important factor.
It will develop from roughly the age of 40 and becomes pronounced after 50.
Another phenomenon that interferes with the lookout function is so-called night
myopia (near-sightedness), which is maintained as long as the eyes are night adapted.
Far-sighted persons may for practical reasons therefore see better without corrective
glasses, whereas a reason with normal sight will improve night vision with a corrective
lens of roughly 1.5 diopters. Jayle et al. (1959) found that when night myopia was
corrected the night vision threshold was improved by 50%.
11.5.4 Uniform Field Difficulty
Apart from the increasing boredom of prolonged observat ion of a uniform field such as
calm sea, open sky or fog, there is also a physiological cost. This may lead to blanking out
after 10–20 minutes. It results in redu ced motor coordination and ability to maintain
balance. Another related phenomenon is empty field myopia, which means that the eye
accommodation is in a constant state of fluctuation due to the lack of objects to focus on.
This problem may be solved by looking away at objects like masts or oth er objects on the
vessel every 5 minutes.
11.5.5 Visual Illusions
Vision is the main source of information for the watchkeeping personnel. In light of this,
it is important to keep in mind that visual illusion is extremely common. Some well-known
factors are:
. Refraction
. Fog and haze
. Texture
. Autokinesis
11.5 SENSING AND OBSER VA TION 32 1
Refraction is the phenomenon whereby the direction of the light is broken by passing
through different media. During rain, water may build up on the windows of the bridge
and result in refraction. This results in a misinterpretation of the relative direction
(bearing) of other vessels or objects.
In fog and haze, objects app ear to be smaller than they actually are and therefore seem
to be farther away. This may put the vessel at risk in approaching situations by reducing
the time and distance to stop or change heading. A compounding factor is also the
tendency to underestimate the relative speed of other objects under marginal visual
conditions. The texture of an object may be a clue about its distance from the vessel.
Fine detail indicates a vessel nearby whereas a diffuse appearance indicates an object
further away. For that reason unusual objects may be misjudged with respect to their
distance.
The visual illusion of autokinesis can happen at night when you look continuously at a
single light against an otherwise dark background. The light will apparently move,
sometimes in an oscillating fashion. Both light markers and lanterns of other vessels might
be the source of this phenomenon.
11.5.6 Radar Operation and Vigilance
Radar operation may be a quite demanding task for various reasons. One situation is
navigation in coastal waters with heavy traffic where the operator has to monitor many
targets and assess collision risk. However, owing to the invention of the ARPA device
(Automatic Radar Plotting Aid), this task has been substantial ly eased. A quite different
challenge is to maintain vigilance during radar watch in open waters with little or no traffic
for longer periods. Mackworth (1950) has defined vigilance as the observer ’s readiness to
detect infrequent, aperiodic, small changes in the environment.
It has been well established that the detection of radar targets may fail for a number of
reasons:
. The signal may be weak
. The radar target is veiled by signal noise
. No warning
. Increased boredom due to the monotony of the task
. Lack of rest pauses
The fact that performance deteriorates with tim e is expressed in the so-called ‘Mariners
Law’: ‘Maximum vigilance can be maintained for a period of about 30 minutes – after
this the performance deteriorates sharply’ (Elliott, 1960). This observation has later
been stated in a more precise form by Teichner (1972), who found that the probability of
detecting a visual signal is a function of:
. The initial probability of detection
. The duration of the watch
. Whether the detection demands continuing adjustment of eye focus or not.
322 CHAPTER 11 HU MAN FA CT ORS
Two of these factors are illustrated in Figure 11.5. This shows that, given a low initial
detection frequency, there is insignificant deterioration with time, whereas a high initial
detection deteriorates sharply during the first 30 minutes. This phenomenon is ascribed
to the fact that the first situation reflects a demanding task requiring high concentration,
whereas the other is less demanding and therefore soon results in loss of concentration.
11. 5 .7 He av y Me als
Madison (1974) has summarized some of the findings of the effect of meals on human
performance. The main and perhaps not surprising conclusion is that the meals should be
well balanced. Experiments and questionnaire studies further indicate that:
. High carbohydrate meals have soporific qualities and give poorer performance.
However, the subjective discomfort was lower.
. High fat meals seems to give a short-term gain in performance
A complicating factor is that individual preferences with respect to meals may have
a major impact on performance and thereby preclude the value of the general findi ngs
cited above.
11. 6 PERCEP TION AND DECIS ION - MAKING
By perception is meant basically the task of making sense of what you detect and observe.
For observation, there are a number of sources of error. Typical questions are: do I see a
vessel or a stationary object, is the light from a marker or another vessel, or is the object
moving away from or towards me? Perception involves:
. Understanding the signal, applying meaningful concepts
. Relating visual input to known pictures
. Assessi ng movement relative to mental, dynamic models
. Giving priority to alternative information sources
Figure 11.5. Initial detection probability versus time on watch for different initial detection probabilities.
(Source: Teichner, 1 972. )
11. 6 PERCEPTI ON AND D ECIS ION -MAK IN G 323
Problems may arise due to the fact that humans apply sequential processing (one thing
at the time), input in short-term memory is forgotten, or mental overload occurs.
The performance of human information processing rests very much also on correct
decision-making: how to assess the situation on the basis of the perceived information
and how to respond in light of the mission plan and the detected obstacles or constraints.
Both perception and decision-making may be flawed by what we call false hypothesis
and habits.
11.6.1 False Hypothesis
A rational perception of what we sense can be influenced by the so-called false hypothesis
phenomenon, or what we may term anticipation or illusions. It appears in different ways
or by different mechanisms:
. A tendency to take in limited information an d assume the rest.
. High expectancy: long experience that things happen in a certain pattern makes you
‘see’ things regardless of the actual stimulus.
. Hypothesis as a defence: interprets things in way that reduce stress or anxiety (not
facing reality).
. After a demanding work period, the concentration or vigilance drops and makes you
more vulnerable to error.
11.6.2 Habit
Safe operation of a vessel is to a considerable degree based on sound habits or, essentially,
behaviours. The objective of training is to establish habitual actions or skills. Through
acquired experience the repertoire of skills is further developed. A negative aspect of time
or age is that physical fitness reaches a peak in the early years. A classic question is
therefore whether the physical superiority of the young is beaten by the greater experience
of the mature crew member. There may, however, be a certain risk with habits under
changing conditions:
. Habitual responses may be inappropriate due to change in the dynamic characteristics
of the vessel, another propulsive system, and so on.
. ‘Cannot make myself forget’: new respo nse patterns have to be trained, but under
stress or with focus elsewhere one returns to earlier habits.
. The problem may be overcome by ‘over-learning’ of critical responses.
. But as new responses and improved skills are developed, the ability to stay calm is also
improved.
11.6.3 End-Spurt Effect
It has been shown that for tasks demanding vigilance, knowledge about the remain-
ing duration of the assignment has an effect on the performance of the operator.
324 CHAPTER 11 HU MAN FA CT ORS
In experiments, subjects knowing how long the time they had to stay on vigil performed
superior to subjects not knowing.
1 1.7 PHYSICAL WORK ENVIRO NMENT
Following physical work, climate factors are thought to be relevant for the performance
of the crew:
. Thermal climate
. Noise
. Vibratio n
. Illumination
These factors were studied in depth by Iverga
˚
rd et al. (1978) on Swedish vessels. The
results wi ll be summarized in this section, but it should be kept in mind that the findings
are not necessarily representative of conditions today.
11. 7.1 T h e r ma l Clim at e
The thermal climate is a function of a number of factors: temperature, humidity, air speed
(circulation) and heat radiation. The subjectively exp erienced climate must further be seen
in relation to the work load and clothing of the individual.
The thermal climate represents a challenge on a vessel for the following reasons:
. It operates in all climate zones
. The conditions vary highly in different sections of the vessel
. The crew members are exposed to different thermal stress
. The crew can be subject to extreme thermal loads in emergency situations
In discussing thermal stress it is necessary to make a distinction between the different
sections of the vessel. Table 11.4 gives a summary of the problems related to thermal
control in the various sections of a vessel.
11 .7.2 Temperature and Vigi lance
The experience of well-being and task performance is a function of the climate. The body
will be in thermal balance if the heat produced by metabolism equates the heat lost
through evaporation, radiation, convection and by work accomplished. As evaporative
heat loss is a key source during most conditions, humidity also plays a key role. In practi ce
we therefore apply the effective temperature (ET) concept, which is a function of
both dry-bulb temperature and relative humidity. The body is in balance with respect
to heat loss at a dry-bulb temperature of 25
C and a relative humidity of 50%
(McCormick, 1976).
11 .7 PHYSICAL WORK ENVIRONMENT 325