23.03.2017 Views

wilamowski-b-m-irwin-j-d-industrial-communication-systems-2011

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Functional Safety 21-9<br />

TABLE 21.4<br />

Safety-Related Communication Systems<br />

Compliant<br />

to Standard<br />

Safe Node Architecture<br />

Safety-Related<br />

Message Format<br />

Safety<br />

Firmware in<br />

the ISO/OSI<br />

Model<br />

CANopen safety IEC61508, SIL 3 Two channel (1oo2) NO—standard<br />

message sent twice;<br />

second time inverted<br />

DeviceNet safety IEC61508, SIL 3 Two channel (2× 1oo1, 1oo2) YES—embedded into<br />

standard protocol<br />

payload field<br />

PROFIsafe IEC61508, SIL 3 Single channel YES—embedded into<br />

standard protocol<br />

payload field<br />

SafetyBUS p EN 954-1, Cat. 4 Two channel (1oo2) YES—embedded into<br />

standard protocol<br />

payload field<br />

SafetyLon IEC61508, SIL 3 Two channel (1oo2) YES—embedded into<br />

standard protocol<br />

payload field<br />

Integrated<br />

into layer 7<br />

Safety layer<br />

above layer 7<br />

Safety layer<br />

above layer 7<br />

Safety layer<br />

above layer 7<br />

Safety layer<br />

above layer 7<br />

non-safe and safe <strong>communication</strong> <strong>systems</strong>. Another reason is that the protocol stack of <strong>industrial</strong> <strong>communication</strong><br />

<strong>systems</strong> is very often standardized and therefore should not to be altered. Table 21.4 gives an<br />

overview of the safety features of some safety-related <strong>communication</strong> <strong>systems</strong>.<br />

As already mentioned, the main goal of safety is to avoid systematic faults and detect stochastic faults<br />

leading to hazards such as a corruption of a message. Safety measures like a CRC are required to mitigate<br />

the risk resulting from the hazards. Typically, standard <strong>communication</strong> <strong>systems</strong> already include<br />

some measures to detect faults (e.g., a CRC or cross-parity mechanism to ensure data integrity). During<br />

safety consideration, there are two approaches of how to treat the standard <strong>communication</strong> system:<br />

either take the available fault detection measures into account and add more measures to finally reach<br />

the target safety integrity level. Or the standard <strong>communication</strong> system with its fault detection measures<br />

is neglected and the system is treated as a black box referred to as black-channel concept.<br />

Safety of control area network (CAN) relies on the fault detection mechanism of standard CAN.<br />

Additionally, to increase integrity and reduce residual failure, probability messages are sent twice and<br />

the second one is inverted. The two standard messages comprise a single safety-related message. To<br />

distinguish safety- and non-safety-related messages, the identifier range in the standard message was<br />

extended. Finally, watchdog functionality is integrated to detect if a safety-related message part or a<br />

complete safety-related message was lost. CANopen Safety meets safety requirements of SIL3 given by<br />

the international standard IEC61508. Since the protocol enhancement was not sufficient to meet the<br />

desired safety level, in addition, a hardware architecture with two channels was chosen.<br />

ProfiSafe or SafetyLon are based on the black-channel concept as shown in Figure 21.5. The network<br />

and the standard protocol, and the standard hardware component (non-safety-related part) of the node<br />

are treated as a black box. Node A inserts a message into the black-channel and Node B receives it. The<br />

idea is that safety measures are required to detect all faults that are possible in the black-channel without<br />

paying attention to existing measures. Typical faults are systematic faults on the network or faults in the<br />

standard hardware component. Any measures of the standard protocol to detect faults like parity bits to<br />

identify message corruption are not considered. The standard protocol is just a means to transport data<br />

somehow from Node A to Node B.<br />

The gray part as shown in Figure 21.5 is the safety-related part. It comprises the safety-related hardware<br />

and firmware. Both are used to increase the safety integrity and consequently reduce the level of<br />

risk to a tolerable target level.<br />

© <strong>2011</strong> by Taylor and Francis Group, LLC

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!