Improving human performance in practice – lectures
Week 1: work and task analysis
Initial focus of ACP engineering and human factors/cognitive ergonomics in for example
military and aviation.
Cognition: Current fields:
Selective attention Knowledge work
Vigilance Creativity and sports
Interference Industry
Capacity High tech and robotics
Identification Education
Decision making Traffic
Government
Task analysis: looking at a non-optimal situation, find the cause and give advice
Drunk car crash advice: don’t drink and drive
However, other possible factors: distraction, fatigue, insufficient training, road design,
fogged windshield interactions between causes.
Resilience to imperfection: aim to ensure that imperfections in people and situations do not
lead to disaster.
Hindsight bias cannot prevent problems
Towards a system analysis (team is more than the sum of its parts)
,Task analysis (light) cognitive psychology allows you to already identify limiting factors in
a not too complicated job:
Situation awareness
Attention/concentration
Working memory load
Knowledge
Availability of information
Week 2 – Safety
Past to present
Human error human performance variability
Motivation training to not pass a red signal analysis of why a signal is being ignored
Railway accidents high availability bias
The “Swiss cheese model”:
Between an accident and a cause are different barriers
An accident occurs when the “holes” in barriers line up
Solution: decrease the size of the holes or add more barriers
The arrow represents the human factor: Why, What and when.
Very often, there is no malicious intent and things “happen”
Unintentional behavior:
Slips commission
Lapses omission increasing problem
Intentional behavior:
Mistakes strong but wrong
Violations breaking rules
, SPAD signal passed at danger from minimal to grave passings
Target was not realized, but there was a significant decrease in SPADS
Variation exist and does not have to be entirely eliminated, however we could implement
thresholds to avoid too high or too low result.
Variation leaves to improvement and competition
Utrecht central station: Opportunities for change: less crossings and more information
processing how would this increase safety?
Unfortunately can’t be just “tried” and has a high cost/high risk nature
The outcome can only be explained after the experiment
How to reduce the risk of SPADS Risk is exposure X probability X effect:
Exposure: influence the number of signals/crossings flexibility lost
Probability (behavior): given that there is a red signal, influence the probability to
passing that signal:
- Using simulators (waiting for a SPAD is too time-consuming)
- Proxy indicator max breaking capacity to SPAD
Effect: Reduce the consequences
Innovative way of improving safety:
There were no indications that negative mBtSPAD were caused by extraordinary
factors
Negative SPADS were extreme outcomes of the accepted way of working initially
not safety related
With a focus on optimizing the primary process, not only safety but also punctuality
and energy consumption.
To what extent is there unnecessary variation in the way organizations operate
Conclusions:
Two approaches that are complementary to each other:
- Eliminate causes of incidents
- Eliminate variation. Change the normal way of operating in such a way that
extreme outcomes are less frequent
Positive surprises are as worrying as negative ones
Safety is a control issue and implies predictability of performance in all aspects
Look for sources for variation in human performance
Week 3 – Education
Predictors of school success 70% of variance depends on gender, SES and IQ
14 years of education to influence the other 30%
What is the goal of school? Is school the optimal mean to achieve the goal applies also to
company training.
Thorndike transfer is only possible from the trained activity to identical elements.
Week 1: work and task analysis
Initial focus of ACP engineering and human factors/cognitive ergonomics in for example
military and aviation.
Cognition: Current fields:
Selective attention Knowledge work
Vigilance Creativity and sports
Interference Industry
Capacity High tech and robotics
Identification Education
Decision making Traffic
Government
Task analysis: looking at a non-optimal situation, find the cause and give advice
Drunk car crash advice: don’t drink and drive
However, other possible factors: distraction, fatigue, insufficient training, road design,
fogged windshield interactions between causes.
Resilience to imperfection: aim to ensure that imperfections in people and situations do not
lead to disaster.
Hindsight bias cannot prevent problems
Towards a system analysis (team is more than the sum of its parts)
,Task analysis (light) cognitive psychology allows you to already identify limiting factors in
a not too complicated job:
Situation awareness
Attention/concentration
Working memory load
Knowledge
Availability of information
Week 2 – Safety
Past to present
Human error human performance variability
Motivation training to not pass a red signal analysis of why a signal is being ignored
Railway accidents high availability bias
The “Swiss cheese model”:
Between an accident and a cause are different barriers
An accident occurs when the “holes” in barriers line up
Solution: decrease the size of the holes or add more barriers
The arrow represents the human factor: Why, What and when.
Very often, there is no malicious intent and things “happen”
Unintentional behavior:
Slips commission
Lapses omission increasing problem
Intentional behavior:
Mistakes strong but wrong
Violations breaking rules
, SPAD signal passed at danger from minimal to grave passings
Target was not realized, but there was a significant decrease in SPADS
Variation exist and does not have to be entirely eliminated, however we could implement
thresholds to avoid too high or too low result.
Variation leaves to improvement and competition
Utrecht central station: Opportunities for change: less crossings and more information
processing how would this increase safety?
Unfortunately can’t be just “tried” and has a high cost/high risk nature
The outcome can only be explained after the experiment
How to reduce the risk of SPADS Risk is exposure X probability X effect:
Exposure: influence the number of signals/crossings flexibility lost
Probability (behavior): given that there is a red signal, influence the probability to
passing that signal:
- Using simulators (waiting for a SPAD is too time-consuming)
- Proxy indicator max breaking capacity to SPAD
Effect: Reduce the consequences
Innovative way of improving safety:
There were no indications that negative mBtSPAD were caused by extraordinary
factors
Negative SPADS were extreme outcomes of the accepted way of working initially
not safety related
With a focus on optimizing the primary process, not only safety but also punctuality
and energy consumption.
To what extent is there unnecessary variation in the way organizations operate
Conclusions:
Two approaches that are complementary to each other:
- Eliminate causes of incidents
- Eliminate variation. Change the normal way of operating in such a way that
extreme outcomes are less frequent
Positive surprises are as worrying as negative ones
Safety is a control issue and implies predictability of performance in all aspects
Look for sources for variation in human performance
Week 3 – Education
Predictors of school success 70% of variance depends on gender, SES and IQ
14 years of education to influence the other 30%
What is the goal of school? Is school the optimal mean to achieve the goal applies also to
company training.
Thorndike transfer is only possible from the trained activity to identical elements.
