4.1 Safety Mechanisms Grouped by Technique References

The provided safety mechanisms grouped under the ISO 26262-5:2018 technique references are provided below.

Table 4-1. Analog and Digital I/O (ISO 26262-5:2018, Table D.5)
Safety MechanismISO26262-5:2018 Technique ReferenceDiagnostic CoverageSafety MechanismSafety Measure
Failure detection (1) by online monitoringD.2.1.1LowSoC online monitoring of the device can be achieved by periodically reading the register map via SPI or I2C.SPI or I2C
Failure detection (2) by online monitoringD.2.1.1LowSoC online monitoring of the device can be achieved by periodically sensing the state of the GPIO pin or GPO pins.GPIO, GPO
Test patternD.2.4.1HighN/AN/A
Code protection for digital I/OD.2.4.2MediumN/AN/A
Multichannel parallel outputD.2.4.3HighN/AN/A
Monitored outputsD.2.4.4High

Any internal status bit can be followed by the GPIO pin or GPO pins:

System PLL loss-of-lock.

Per-reference single-cycle monitor, precise frequency monitor, and other monitors.

DPLL loss-of-lock, holdover, state, selected ref, and other status.

GPIO, GPO
Input comparison votingD.2.4.5HighN/AN/A
Table 4-2. Communication Bus Serial/Parallel
Safety MechanismISO 26262-5:2018 Technique ReferenceDiagnostic CoverageSafety MechanismSafety Measure
One-bit hardware redundancyD.2.5.1LowN/AN/A
Multi-bit hardware redundancyD.2.5.2MediumN/AN/A
Read back of sent messageD.2.5.9MediumValue previously written over SPI or I2C can be read back from the register mapSPI or I2C
Complete hardware redundancyD.2.5.3HighN/AN/A
Inspection using test patternsD.2.5.4HighN/AN/A
Transmission redundancD.2.5.5MediumN/AN/A
Information redundancyD.2.5.6MediumN/AN/A
Frame counterD.2.5.7MediumN/AN/A
Timeout monitoringD.2.5.8MediumN/AN/A
Combination of information redundancy, frame counter, and timeout monitoringD.2.5.6 D.2.5.7 D.2.5.8HighN/AN/A
Table 4-3. Program Sequence Monitoring/Clock (ISO 26262-5:2018, Table D.8)
Safety MechanismISO 26262-5:2018 Technique ReferenceDiagnostic CoverageSafety MechanismSafety Measure
Watchdog with separate time base without time windowD.2.7.1LowN/AN/A
Watchdog with separate time base and time windowD.2.7.2MediumN/AN/A
Logical monitoring of program sequenceD.2.7.3MediumUse SPI or I2C access to read the current DPLL state periodically.State Monitor
Combination of temporal and logical monitoring of program sequenceD.2.7.4HighN/AN/A
Combnation of temporal and logical monitoring of program sequences with time dependencyD.2.7.5HighN/AN/A
Table 4-4. Sensors (ISO 26262-5:2018, Table D.9)
Safety MechanismISO 26262-5:2018 Technique ReferenceDiagnostic CoverageSafety MechanismSafety Measure
Failure detection by online monitoringD.2.1.1LowN/AN/A
Test patternD.2.4.1HighN/AN/A
Input comparison/ votingD.2.4.5HighN/AN/A
Sensor valid rangeD.2.8.1LowN/AN/A
Sensor correlationD.2.8.2HighN/AN/A
Sensor rationality checkD.2.8.3MediumN/AN/A
Table 4-5. Non-volatile Memory (ISO 26262-5:2018, Table 32)
Safety MechanismISO 26262-5:2018 Technique ReferenceDiagnostic CoverageSafety MechanismSafety Measure
Parity Bit5.1.13.6LowN/AN/A
Memory monitoring using error detection- correction codes5.1.13.1HighN/AN/A
Modified checksum5.1.13.2LowAt power-up data in internal flash memory is loaded to RAM. A checksum is used to ensure the transfer is accurate.Boot Status
Memory signature5.1.13.3HighN/AN/A
Block replication5.1.13.4HighInternal flash memory can have two copies of all data.Golden and Upgrade copies, Boot Golden command