Information Technology Reference
In-Depth Information
Thus, it seems good advice to use both approaches (where possible) and to
use one method as a plausibility check for the results of the other method.
The optimal approach would have to be a fixed assignment from accident
scenarios to severity classes so that the variations in expert judgements are
minimised. However, already from Table 2, it is apparent that classes P-F
and P-S are quite narrow and may be hard to be distinguished from their
neighbours. From this observation, the consolidated severity categories shown
in Table 4 result.
Tabl e 4 . Consolidated severity categories
ID Combinations
FWI range
Typical FWI
E
M-F
2FWI
5
D
P-F, M-S
0.2 FWI<2
1
C
P-S, M-L
0.02 FWI<0.2
0.1
B
P-L
0.01 FWI<0.02
0.01
A
-
FWI<0.01
n. a.
In Table 4, simple letters are used to denote the severity categories and
unambiguous, sometimes also misleading, verbal descriptions.
Table 5 shows an example for a simple classification of accidents with
respect to their severity. With such a table, an unambiguous and easy-to-use
classification would be possible. Similar classification tables are already in
use in the car industry, e.g. the new draft ISO 26262 safety standard (2010).
Tabl e 5 . Examples of accident classification
ID
Derailment...
Collision...
Impact
Personal accidents
E
Passenger train
at high speed
Passenger train on
a main line
-
D
Passenger train
at
At a level crossing Train with
work gang
Passenger falling
out of a train at
high speed
medium
speed
C
Passenger train
at low speed
Passenger falling
out of a train at
low
speed
or
at
stop
B
In shunting op-
eration
Train into
buffer at low
speed
Passenger hit by
a door Passenger
falling during em-
barkment
A
 
Search WWH ::




Custom Search