What Is Redundancy? How Backup Components Improve Robot Reliability.

What Is Redundancy?

Redundancy is the use of additional components, systems, or methods to support a function when one element fails or becomes unreliable.

In robotics, redundancy can help maintain important capabilities or support safety-related fault detection.

For example, an autonomous machine may use more than one sensor to monitor a critical condition.

If one sensor fails, another measurement may help the system detect the problem or continue operating within defined limits.

Key Takeaways

• Redundancy adds alternative ways to perform or monitor a function.
• It can improve reliability and support fault detection or fault tolerance.
• Additional components do not automatically improve safety if they share common failure causes.

How Does Redundancy Work?

Imagine an industrial robot with a safety-related monitoring function.

The system may use two independent signal channels.

Under normal conditions, both provide information consistent with the expected operating state.

If one channel produces a contradictory signal, the system can recognize a possible fault.

The controller then initiates the response required by its safety design.

This example shows redundancy supporting fault detection.

Other systems use redundancy to maintain operation.

For example, a computing platform may use backup processing hardware that can assume certain functions after a failure.

The purpose and architecture depend on the application.

Hardware vs. Software Redundancy

Hardware redundancy: Uses additional physical components, such as sensors, processors, or power supplies.

Software redundancy: Uses alternative computations, algorithms, or implementations to provide or check a function.

For example, a robot may estimate movement using wheel encoders and an inertial sensor.

Those measurements can provide complementary information.

However, using several sensors does not automatically establish that the system has a fully redundant safety architecture.

The design must account for how the measurements depend on one another.

Redundancy vs. Fault Tolerance

Redundancy describes the presence of additional resources or methods.

Fault tolerance describes the ability to maintain specified functionality despite certain faults.

Redundancy is one way to support fault tolerance, but the terms are not interchangeable.

A robot may contain two sensors yet still stop after one fails.

Another system may continue operating using the remaining measurement.

The actual behavior depends on its software and safety requirements.

What Is a Common-Cause Failure?

A common-cause failure affects multiple components because they share a vulnerability or dependency.

Consider two sensors connected to the same power supply.

If the power supply fails, both sensors may stop functioning.

Similarly, two redundant computers may run the same defective software.

The additional hardware does not eliminate that shared weakness.

Engineers therefore examine independence, diagnostic coverage, and common failure scenarios when designing redundancy.

Is More Redundancy Always Better?

Not necessarily.

Additional components introduce cost, weight, maintenance, and engineering complexity.

A small robot may not need duplicate versions of every sensor.

Redundancy should be selected according to the importance of the function and the consequences of failure.

Frequently Asked Questions

Does redundancy prevent every robot shutdown?

No. Shared dependencies, software faults, and unanticipated failures can still cause interruptions.

Do safety-related functions always use two sensors?

No. The appropriate architecture depends on the required safety performance and application.

The Bottom Line

Redundancy provides alternative components or methods that can improve reliability and fault handling. Its effectiveness depends on how independently those alternatives operate and what happens when a failure is detected.