Reliability block diagrams · Interactive demo
Reliability block diagram demo.
See how redundancy changes the reliability of a simple sensor, controller and actuator chain. Change the architecture and failure rates, or select a block to fail it.
How long the system must keep working, in operating hours. Reliability falls as this grows, so a result only means something next to its mission time.
The probability, from 0 to 100%, that the chain still works at the end of the mission time.
The probability that it has failed by then. With very high reliability, this number is easier to compare.
Actuator is the stage most likely to have failed by the end of the mission. Try adding redundancy there, then compare the result.
How your result is worked out
These lines update with your values. Each block has a constant failure rate, so a block's reliability is R = e−λt.
| Stage | Block reliability, R | Stage formula | Stage reliability | Stage unreliability |
|---|---|---|---|---|
| Sensor | e^(−150 FIT × 10⁻⁹ × 10000 h) = 0.998501 | 3R² − 2R³ | 0.99999327 | 6.73e-6 |
| Controller | e^(−300 FIT × 10⁻⁹ × 10000 h) = 0.997004 | 1 − (1 − R)² | 0.99999103 | 8.97e-6 |
| Actuator | e^(−50 FIT × 10⁻⁹ × 10000 h) = 0.999500 | R | 0.99950012 | 5.00e-4 |
Series chain: multiply the three stage reliabilities to get R(t) = 99.9484%.
How to read a reliability block diagram
An RBD shows what must work, not what can fail. Follow the green path from left to right: the system works as long as one complete path of working blocks remains. ISO 26262-9:2018, Clause 8.2 lists RBDs among the quantitative safety analysis methods. The conventions here follow IEC 61508-6:2010, Annex B.
Single block
Blocks in a line are in series. One failure breaks the path, so the chain depends most on its weakest stage.
1oo2
One out of two must work. Both channels must fail before the stage fails.
2oo3
Two out of three must work. Any two failures make the stage fail, but one wrong channel can be outvoted.
This demo assumes independent, non-repairable blocks. Real channels can share a supply, clock or design error, and diagnostics and repair also change the result. An RBD reliability value is therefore not a PMHF. To estimate that, use the hardware metrics calculator.
This demo and the full simulator
| Feature | This demo | Full simulator |
|---|---|---|
| Diagram | One fixed sensor, controller and actuator chain | Your own blocks, with nested groups |
| Redundancy | Single block, 1oo2 or 2oo3 per stage | Series, parallel and any k-out-of-n group |
| Results | R(t), unreliability and each stage’s relative chance of failure | Reliability curves over time, MTTF and importance measures |
| Fault tree | Not included | Equivalent fault tree and minimal cut sets |
| Examples | One default chain | Worked automotive examples, such as a brake-by-wire controller |
Go further
Model your own system in the full RBD simulator.
The full RBD simulator is included with a free account.