Drones are driven by an interplay of lots of technological and engineering breakthroughs. These make for affordable production and safe, efficient flight.
One of these breakthroughs is that of redundant systems. Many top drones are using them but the truth is what many people are sat wondering ‘what on earth is a redundant system?!’
Don’t worry, we’re here to explain.
What Are Redundant Systems in Engineering?
Redundancy is an engineering concept that involves the duplication of critical components or functions within a system, with the aim of increasing reliability.
Redundancy measures usually take the form of a backup or fail safe, or measures that seek to improve the actual system performance, like is common in GNSS (Global Navigation Satellite System) receivers and multi-threaded computer processing.
Sometimes a system is too critical to merely duplicate its supporting components. So engineers triple them in a process known as triplication.
Many safety-critical systems, such as fly-by-wire and aircraft hydraulic systems, as well as some parts of aircraft control systems, are triplicated in a process called triple modular redundancy (TMR).
Lives depend on it.
In redundant systems, an error in one component may be kept supported by the backup. In triply redundant systems, the system is made of three sub components and all three must fail before the system fails.
Each sub-component is made to the utmost quality to reduce the likelihood that it will ever fail. More to that, each sub-component is expected to fail independently. Hence, the probability of all three failing is calculated to be extremely small, and is often outweighed by other risk factors like human error.
Redundancy is also sometimes known as “majority voting system” or “voting logic” and sometimes it could lead to a more complex system than otherwise expected.
Some of the problems that arise with redundant systems could be as simple and avoidable as neglect of human duty due to complacency and over stressing of the operating mechanisms.
Functions of Redundancy in a System
To understand the function of redundancy, you have to become acquainted with the two variants of redundancy which are passive redundancy and active redundancy.
Both of these prevent performance decline from exceeding specification limits without the need for human intervention.
Passive redundancy uses excess capacity to reduce the resulting impact of a component failing. One of the best examples of passive redundancy in action is the extra strength of cabling and struts used in bridges.
This extra strength makes it possible for a bridge to keep standing even after some structural components have fallen away. We call the extra strength hidden in the design the margin of safety.
Other good examples of passive redundancy are provided by your eyes and ears.
Vision loss in one eye does not cause blindness, although your perception of depth is impaired. Likewise, the loss of hearing in one ear does not cause total deafness, even though you may lose your sense of directionality.
To cut things short, you can expect passive redundancy to result in performance decline when a limited number of failures occur.
Active redundancy, on the other hand, eliminates a decline in performance by monitoring the performance of individual devices. And this monitoring takes the form of voting logic.
The voting logic consists of electronic switching that automatically reconfigures the components. A good example of active redundancy is error detection and the Global Positioning System (GPS).
Another example of active redundancy is provided by electrical power distribution systems. You have several power lines connecting each generating facility to customers.
Each power line includes monitors that detect any overload and also circuit breakers. The combination of power lines makes for excess energy capacity. But the circuit breakers will disconnect any power line on which an overload is detected. After this happens, power is redistributed across the remaining lines.
Active redundancy systems work by way of voting logic. And this is a system that are systems that monitor performance in order to determine how to reconfigure individual components so that operation continues without violating limits specified within the overall system.
It’s more common to see these used in computers, but they can also be set up in systems that are not made of computers specifically.
Electrical systems, for example, use active redundancy to adjust the output of each generating facility when other facilities are suddenly lost. And that’s how blackouts are prevented during major events like earthquakes.
In computer systems, a simple form of voting logic is often used that involves two components: a primary and an alternate.
These run on similar software, but the alternate does nothing during normal smooth running operations.
The primary monitors itself and periodically sends an activity message to the alternate — almost like a notification that all is well and good.
Things change when the primary detects a fault, however. At that point, all output from the primary stops, including the activity message that is due for the alternate.
The alternate is then activated to start working once this happens, and it takes over from the primary when the activity messages cease.
An alternate form of voting logic, and a more reliable one, involves an odd number of devices — usually three or more. All perform identical functions and the outputs are compared by voting logic.
The voting logic establishes a majority when there is a disagreement in output, and the majority devices will deactivate the output from the other devices that disagree.
In this way, a single fault does not interrupt normal operation.
This is the technique used for avionics systems such those of a space shuttle or drone.
What is the Importance of Redundant Systems?
As you can imagine, redundancy in a system is crucial for the smooth running of the system.
In certain devices, like bridges, power lines, and aircraft it could go a long way towards ensuring safety of lives and equipment.
It is no surprise therefore that drone manufacturers go all out in incorporating redundancy systems in their devices.
One of the key reasons for this is safety, but there are also many other reasons why drones are created with many components that do the same job, despite the resulting extra weight.
How Do Drones Use Redundant Systems?
Drone systems usually incorporate fail proof mechanisms using automations like “Go Home” or emergency landing, which kick in when there is a system failure or a lack of battery life.
But what happens when there is a system failure that affects these fail proof systems?
It’s scenarios like this which warrant the need for redundant systems like autopilot that are able to lower the possibility of outright system failure to a minimum.
Drone manufacturers incorporate multiple sensors that do the same job in their systems in order to prevent drones from immediately falling when, say, a motor fails to rotate a propeller in mid-flight.
And yes, this does cause the drone to be heavier than it would otherwise be without redundancy. But redundant stabilization systems improve drone performance and can boost sensor accuracy many fold.
To best illustrate this, we need to take a look at how drones work.
The flight controlling component of a drone is what receives signals from you, the pilot (or the autopilot), who makes use of GPS to take charge of the flight path.
Drones also have inertia measurements units (IMU) that help in judging their posture during flight.
For a safe flight, the flight controlling component needs to know so many things in order to coordinate proceedings; stuff like destination, altitude, GPS coordinates and IMU signals.
All the sensors that provide these details simply cannot lag in accuracy or everything would come apart in midair.
Manufacturers, thus, ensure unfailing accuracy by building redundancy into their products.
For example, the mainline Phantom 4 has a double compass system that detects information from IMUs and the drone’s direction.
The Phantom 4 is also equipped to receive information from up to 24 satellites aligned with either GPS or Russia’s Global Navigation Satellite System (GNSS). This means the system holds no matter which part of the globe you fly your drone in.
The very portable Mavic Pro (and even the Phantom 4) is equipped with two camera sensors to detect conditions on the ground, as well as two ultrasonic height-measuring sensors.
Along the same vein, the large industrial-use Matrice 600 rotates itself to avoid falling — even if three of its six propellers drop off. It also mounts up to six batteries to ensure continuous power supply come what may.
Research is also presently being done to produce drones that are designed to continue operating even when one of its propellers fails.
Such a system would require a voting logic mechanism that monitors the propellers and recognizes when one propeller stops working, so that it can immediately adjust the remaining rotors as appropriate.
Researchers involved also intend to create a system that is capable of detecting signs of impending malfunction so that adjustments can be made before anything stops working.
Other flight researchers like those at the Autonomous Control Systems Laboratory in Chiba, Japan are working on drones that come with parachutes.
Once a sensor recognizes a drone’s tilt and falling speed, it would be able to automatically open the parachute in a tenth of a second — almost like an airbag for drones. The parachute would also be deployed when the drone’s batteries fail.
The parachute would be able to guarantee a slow enough descent to ensure a safe landing.
These are just some of the ways in which redundancy is making the difference in your drone and guaranteeing you a safe and seamless flight.
And while it does make for heavier, more power-consuming drones, few pilots would complain when they get to realize how these systems operate to safeguard their hardware.