|
Selection of methods for the correctness of implementation of functional safety requirements
To ensure the correctness of implementation of functional safety requirements the following
types of test are all highly recommended at all ASIL levels
Requirements-based test
Fault injection test
Long term test
User test under real life conditions
Requirements based testing is used to test against functional and non-functional (quality of
service) requirements.
Fault injection test uses special means to introduce faults into the item. This can be done within
the item via a special test interface or specially prepared elements or communication devices. The method is often used
to improve the test coverage of the safety requirements, because during normal operation safety mechanisms are not
invoked
A long term test and a user test under real-life conditions are similar to tests derived from field
experience but use a larger sample size, normal users as testers, and are not bound to prior specified test scenarios,
but performed under real-life conditions during everyday life.
This summary is based upon "Table 15- Correct implement of the functional safety requirements at
the vehicle level" located in ISO 26262, vol 4, clause 8.4.4.2.2
|
Selection of methods for the correctness of functional performance, accuracy and timing of safety mechanisms (A), (B), C, D
To ensure the correctness of functional performance, accuracy and timing of safety mechanisms the
following types of test are required for items of ASIL C or D and recommended for items of ASIL A or B
Performance test
Long term test
User test under real-life conditions
Performance testing is used to verify how the safety mechanisms of the item perform, these tests
cover aspects such as fault tolerant time intervals and vehicle controllability in the presence of faults.
Long term and user test under real life conditions, use a large sample size with normal users as
testers. The users are not bound to pre- determined test scenarios, but the tests are carried out during normal
real-life conditions.
Please refer to "Table 16 - Correct functional performance, accuracy and timing of safety
mechanisms at the vehicle level", located in ISO 26262, vol 4, clause 8.4.4.2.3 for the full details.
|
Selection of methods for the consistency and correctness of implementation of external interfaces (A), (B), C, D
These tests are required for items of ASIL C or D and recommended for items of ASIL A or
B
Please refer to"Table 17- Consistent and correct implmentation of internal and external interfaces
at the vehicle level", located at ISO 26262, vol 4, clause 8.4.4.2.4 to find the recommendations regarding testing
of interfaces. In summary the tests covered are the following:
Test if external interfaces
Test of interaction/communication
Both types of tests are highly recommended for items of ASIL C or D and recommended for items of
ASIL B with an unbiased opinion on its usage for items of ASIL A
Testing of external interfaces at the vehicle level is an examination of the interfaces of the
vehicle for compatibility, These tests are done statically (by validating value ranges, ratings or geometries)
and dynamically (during operation of the whole vehicle).
Communication and interactions tests are carried out on the vehicle at runtime and check the
communication between systems against the functional and non functional requirements.
|
Selection of methods for the effectiveness and failure coverage of safety mechanisms at vehicle level (A), (B), C, D
These tests are required for items of ASIL C or D and recommended for items of ASIL A or B
To ensure the effectiveness of the diagnostic coverage of the hardware fault detection, please refer to "Table 18 -
Effectiveness of a safety mechanism's failure coverage at the vehicle level" located at ISO 26262, vol 4, clause
8.4.4.2.5. this table covers
Fault injection tests
Error Guessing
Tests derived from field experience
All these techniques are recommended for ASILs A or B and highly recommended for ASILs C or D
Fault injection introduces faults into the model at runtime to determine the how the software reacts. This is normally
achieved through a special test interface or hardware. This technique is often employed to improve test coverage of the
safety requirements, this because during normal operation safety mechanisms are not involved
Error guessing employs expert knowledge and data collected from experience or the field. This is used to anticipate
possible errors in the test object.
Tests derived from field experience use the experience and data gathered from the field. Erroneous vehicle behavior or
newly discovered operational situations are analyzed and a set of tests is designed to check the vehicle with respect
to the new findings.
For references to fault models see ISO 26262 vol 5: , Annex D, Table D.1
|
Selection of methods for the level of robustness at the vehicle level (A), (B), C, D
These tests are required for items of ASIL C or D and recommended for items of ASIL A or B
To determine the test required to ensure the Level of robustness of the item, please refer to "Table 19-Level of
robustness at the vehicle level" located in ISO 26262, vol 4, clause 8.4.4.2.6, in summary the tests to be carried out
are
Resource usage test
Stress test
Test for interference resistance and robustness under certain environmental condition
Long term test
All these tests are recommended for ASIL B and highly recommended for ASILs C and D with an unbiased opinion on there usage for items of ASIL
A
At the item level lab cars or prototypes are usually employed to create a dynamic environment for the tests. The
focus of these test are items such as power consumption or bus loading.
Stress tests are used to verify the correct operation of the item under high operational loads or extreme environmental
conditions.
A test for interference resistance and robustness under certain environmental conditions is a special case of stress
testing. This includes EMC and ESD tests.
Long term and user test under real life conditions, use a large sample size with normal users as testers. the users are
not bound to predetermined test scenarios, but the tests are carried out during normal real-life conditions.
|
|