Professional Documents
Culture Documents
Integration of the
Human Factor
a) (b) (c) (d) (e)
Baltasar Gil, Pilar Calvo, Antonio Monrocle, Víctor Irala, Iñaki Aliaga
c) (d)
Renfe Operadora
(a) (b)
ESM Institute of Research into Safety and Human Factors
(e)
Lander Simulation and Training Solutions, S.A.
Keywords: Renfe Operadora, ESM, Institute of Research into Safety and Human Factors, and Lander
K1, Simulation Simulation and Training Solutions, are aware of the new scenario created by railway
K2, Human Error interoperability and the latest ERTMS-ETCS driving systems, as well as safety demands
K2, Assessment coming from European Union Directives, and have detected the need to develop a
K4, Training methodology with simulators capable of identifying and evaluating, in an objective and
automatic way, human errors in train driving through simulation training. Based on that need,
we have created “multipurpose simulators”, which are capable of evaluating human error in
rail safety, while integrating and quantifying Human Factors in the training process.
Corresponding author:
Baltasar Gil de Egea The structure of the teaching methods will be devised to fulfil the training and assessment
Tel.: (34) 985 235934 objectives defined by the instructors of Renfe Operadora. The achievement of these training
Fax.(34) 985 273832 and assessment goals will make it possible to guarantee that drivers are efficiently trained to
e-mail: baltasargil@esm.es perform their task in a professional manner, ensuring that they fulfil the legal requirements in
Address: C/Martínez Cachero, 12 this matter.
bajo33013 Oviedo – Asturias – Spain
All the international research studies agree on the fact and its contribution to risk and it is independent of the
that human error is involved in 80% of accidents and specific circumstances surrounding the event, whereas
incidents, not only in the rail sector, but also in the Specific Values do depend on the specific conditions
aeronautics and the marine sector. leading up to the error in the analysed event.
from a professional to trainees, and the assessment of trainees are able to anticipate the
how the trainees had assimilated such knowledge. consequences even if they have not personally
Nowadays, however, with the use of simulators, trainees experienced negative consequences before,
must prove, not only that they have assimilated thus encouraging the development of risk
theoretical knowledge or that they have a certain number awareness.
of driving hours under their belt, but also that they have
attained a certain degree of competence, that is to say
the ability to combine theoretical and practical knowledge • Developing specific exercises aimed at
to sort out a given difficult situation –both in highly improving risk perception, with a view to
automated situations and in degraded operations. This is recycling agents who show a poor
why teaching methods, assessment methods and human understanding of risk perception or who expect
reliability improvement methods need to move forward. to see a particular set of signals only because
they have been exposed to an intense
repetition of a particular signal associated with
In this way, simulation training must bring about a particular traffic configuration.
a change in the situation from which the trainees, whether
they are experienced or not, start their training.
5. Assessment
Improve training by:
Assessment of a training session is based both
• Continuously assessing, on a personal basis, on the objective data provided by the system and on the
the level of competence throughout the assessment previously issued by a panel of expert
learning process. This personalised method instructors, from which all subjectivity has been removed.
usually increases personal motivation. A training tool, such as a simulator, must have a series of
features which are capable of analysing and assessing
the exercises performed in it by the trainees, in a
• Identifying inadequate behaviour under certain
comprehensive, precise, and specific way, subject to
driving conditions and learning gaps. This
objective appraisal. Any simulator should, therefore, be
enables further “a la carte” training, which
able to simplify the instructor’s assessment tasks. This
suits the individual needs of trainees better,
includes predicting and monitoring human errors, as well
while at the same time maximising the use of
as increasing drivers’ reliability.
available resources.
• Identifying reliable behaviour with a view to Objective Assessment is based on statistical data
positively reinforcing it, as a model for other and analysis of the simultaneous occurrence of certain
trainees. variables collected within the parameters of the
performance of an exercise in the simulator. In this way,
Acquire new knowledge and abilities or skills by: allowing the system to obtain in real time the readings of
certain indicators in real time, which then become the
• Implementing changes in the exercises. These criteria on which to assess the trainee’s competences
changes could contemplate, for example, tested in the simulator.
infrequent or unusual traffic conditions, with a
view to highlighting active safety-critical errors
and predicting behavioural limits in tasks The capacity to collect and process data is one of
prone to the generation of human error. the areas in which this type of simulator outperforms other
more traditional training systems, which show a higher
degree of subjectivity. Consequently, we can say that
• Training in situations where the trainee must
simulation training is more objective and complete. We
demonstrate reliable behaviour, especially
can then talk about a true Competence Assessment.
when the task requires him to make a
calculated decision between “safety” and
“production”, involving critical situations under
stress. However, it is very difficult to assess some
factors in situations where the relevant regulations only
recommend that “we proceed with care” or words to that
• Having the instructor come up with effect. In such cases, the instructor must rely on his
“customised” exercises. He can do so by experience as an added-value tool, introducing data into
increasing the level of difficulty according to the system in a quantifiable way, so as not to permit
each individual trainee, by modifying the variations even if the data were to be assessed by a
exercise settings in order to obtain a wider different instructor.
range of situations focused on the same skill,
and by aiming at the consistent definition and
memorising of procedures, through a well- In these cases, the instructor must become
designed and continuously updated teaching aware of the situation beforehand, in order to input the
programme. quantifiable parameter into the system, in real time or at a
later stage. Nevertheless, to facilitate the instructor’s task
Change attitudes by: and level off the criteria of the various instructors, there
are a number of categories at their disposal which allow
• Promoting training in preventive attitudes in them to analyse, in a comprehensive way, the above-
usual or unusual operations, so that the mentioned aspects of driving.
The aim of the project was not to lose the training scope
of traditional simulators, but to improve the cost-
effectiveness and the reachable driver population per
instructor. To achieve that, a specific design of the
Trainee’s Station, Instructor’s Station and Training Centre
Network was developed, manufactured and
commissioned between 2006 and 2008.
Figure 2: Driving desk design
MPS: Trainee’s Station (TS) The level of configurability achieved by this solution of
driving desk fulfilled the requirements of Renfe. Different
One of the biggest issues in the design of a simulation configurations can be seen in Figure 3 and Figure 4,
cab is to maximize the degree of immersion a trainee will demonstrating much greater training possibilities than in
experience inside, so that a great transference to real life a replica cab.
is achieved by means of simulation based training. For
this purpose, a specific cab design was developed, in
combination with a front projection solutions to cover a
wide range of the front field of view of the driver, as
Figure 1 shows.
non-automatic blocks in single and double- • Not only one but several simulations can take
tracked sections, with and without CTC, etc. place at the same time with just one Instructor.
• Logistic problems reduced by installing several
MPS: Training Centre Network Training Centres in different locations so that the
organization of training sessions is easier and
Given the low cost of the simulation cabins (the cost of MUCH CHEAPER
the cab is reduced 10 to 1) and the enhanced automatic • The complexity of the system is decreased:
assessment module of the IS, a new layout is feasible, in • No special needs in client’s facilities to host the
which we forget the traditional one-to-one training and simulators. Normal offices are enough.
come to a new layout of Training Center, in which more • Lower installation costs.
than one cabin can be controlled from a single IS and by • Extremely little maintenance.
a single Instructor, even with all of them working at the • More advanced assessment tools, to raise the
same time. level of delivered training from the point of view
of both, Trainer and Trainee.
The network of Training Centers of Renfe consists of • The cost of one new train to be simulated on a
10 of them, each equipped with a set of Trainee’s MPS is much cheaper than a Full Mission.
Stations, managed from one Instructor’s Station. The • If many trains, ATPs, lines, incidences, etc. are
global layout is as follows: included, the development cost is increased.
However, the final product will be very cost
Training Centres with 3 TS: Madrid (two Centres, see effective.
Figure 7)
Training Centres with 2 TS: Barcelona (two Centres), 7. Findings and Conclusions
Seville, León, Valencia and Bilbao.
Training Centres with 1 TS: Miranda and Santiago All the errors registered are treated by using a
series of Human Error Analysis tools, allowing the
organization to find out which type or types of errors are
safety-critical for each particular operation and to know
whether the errors are consistent or random. The link
between active errors and their underlying factors or
precursors, together with the statistical analysis of the
results, encourages us to move forward with confidence in
the same direction.
of indicators, lamps, HMIs, etc. represented by means of Wilson, J., Norris, B., Clarke, T., and Mills, A. Rail Human
three touch screens in the front desk. Factors. Supporting the Integrated Railway. Ashgate Ed.
2005.
The simulation includes train units and locomotives
(freight and passengers). Simulation of the signaling is
automatic. Malfunctions, incidents and adverse weather Wittingham, R.B., The Blame Machine: why Human Error
conditions can be simulated. The training includes causes Accidents. Ed. Elsevier. 2003.
ordinary and emergency protocols, communications,
regulations... Automatic mistake detection and playback
mode are also available.
8. References
Building a Safe, Interoperable Railway: A Methodological
Guide to Integrating Human Factors. UIC. 2004.
Wilson, J., Norris, B., Clarke, T., and Mills, A. People and
Rail Systems. Human Factors in Road and Rail
Transport. 2007.