Inertial Reference System (IRS)

Inertial Reference System (IRS)


A clear and universally agreed upon definition for the term “Inertial Navigation System (INS)” is not possible. On the one hand the term INS is used as a blanket description for a wide variety of navigation sensors and systems of different design; and on the other hand, it is also used to describe a specific version of these sensors and systems! The term has also changed over the years as the technology has improved. What can be said with confidence is that all these systems work on a similar principle and for the same purpose.

Below is a list of commonly used terms that are used colloquially and interchangeably by pilots (if not by all designers, manufacturers and engineers) to mean very much the same thing, with differences in some of the detail. Where necessary two or more definitions are provided.

Inertial Navigation

a form of “Dead-Reckoning” that relies on accelerometers and gyroscopes to detect acceleration and velocity respectively along 3 perpendicular axes. An approximate 2 or 3 dimensional position can be constantly determined in relation to a known starting point, velocity and orientation.

also a term used to refer to the whole subject: theory, design, technology, and application.

Dead Reckoning

a type of navigation from a known starting point and then by using vector information (direction and speed) against a clock, an estimate of the current position can be made. An INS will calculate a continuous dead-reckoning position. Position accuracy is dependent on the accuracy of vector information and the time since the last “known” position.

Inertial Platform

refers either to the self-referencing gyro-stabilised platform on which the accelerometers are mounted, or

refers to the aircraft, or the neutral aircraft position (pitch, roll and yaw axes), to which a strap-down inertial navigation system is firmly attached.

Inertial Reference System (IRS)

refers to a solid-state unit of three Ring Laser Gyros detecting accelerations in 3 dimensions; they may also contain quartz accelerometers.

Inertial Reference Unit (IRU)

refers to a computer that integrates IRS outputs and provides inertial reference outputs for use by other navigation and flight control systems, including the Flight Management System (FMS).


has traditionally referred to mechanical spinning gyros set within gimbals and frames allowing a platform to be stabilised in space regardless of the motion of the aircraft.

nowadays the term Gyro usually refers to a solid-state Ring Laser Gyro (RLG), which is predominantly used in modern aircraft INS.

several other types of Gyro are available, such as:

  • Hemispherical Resonating Gyros (Wine Glass Gyros), which use hollowed out quartz resonators in which harmonic standing wave rotations are measured.
  • Tuning Fork Gyros use electrically driven quartz tuning forks on a silicon chip; movement will twist the forks and this changes the capacitance between the tines proportional to angular movement, which can be measured.
  • MEMS (MicroElectricalMechanical) Gyro.
  • Vibrating Wheel Gyro.
  • Cylindrical Vibratory Gyro.
  • Piezoelectric Gyro.

Inertial Navigation System (INS)

traditionally (1960’s and 70’s) referred to a self-contained navigation system utilising a gyro-stabilised platform for dead-reckoning, and with a pilot interface allowing a limited number of waypoints to be entered and basic navigation information to be displayed.

then (1980’s) referred to a similar system as above but with the additional capability of feeding inertial data to other aircraft systems such as the Flight Director and Autopilot.

nowadays is predominantly used to describe an Inertial Platform, Inertial Reference System (IRS) and Inertial Reference Unit (IRU); and often used to describe all three of these units (as one system) typically installed on a single aircraft.

INS Position

refers to the latitude and longitude, and altitude, being fed from the IRU to other systems for reference, or as selected for display by a pilot on the Electronic Flight Instrument System (EFIS).

There are many other interchangeable terms in use such as: Inertial Guidance System, Air Data Inertial Reference Units and Systems, Inertial Measurement Unit etc, which can all be used when referring to INS. For the purposes of this Article, the definition of INS (what it is and what it does) that is likely to be most commonly used in the aviation community is as follows:

Regardless of the technology used and the configuration of individual components, an Inertial Navigation System (INS) refers to the Inertial Platform, Reference System and Unit, and associated Inertial Outputs. Its purpose is to provide a continuous dead-reckoning position, altitude and speeds to the FMS, where this data is compared with inputs from other systems, such as ground-based and space-based navigation systems. This other navigation data can be used to update the INS. Data from the INS can also be used for input to flight control and guidance systems.


An INS usually requires an initialisation process that establishes the relationship between the aircraft “frame” (the reference axes) and the geographic reference (position and orientation). This process is called alignment. Alignment usually requires the aircraft to remain stationary for a period of time in order to initialise fully. Alignment times vary depending on the technology in use and the accuracy required from the INS. Alignment will normally require the input of certain data from another system or from a manual entry. Some types of INS can align while “in motion” with the aid of GPS, but performance is then usually reduced or degraded; but it can still provide a useful attitude reference. Accurate alignment is essential if the INS is the sole means of navigation for any length of time without external position updates available.

Alignment can be achieved independent of any external data; this is known as self-alignment. Or, the alignment process can be speeded up with data supplied from a GPS or other systems, and even manual entry.


acceleration due to gravity is always perpendicular to the Earth’s surface and this provides the vertical reference, allowing output of attitude.


the INS can detect Latitude during the alignment, however it will require and accurate Longitude input. Usually the system memory will contain the position where the INS was last shut down, and as long as the aircraft has not been moved this position can be used to speed up the alignment. Once the vertical reference is attained, and the aircraft remains stationary, the only movement to be detected is the Earth’s rotation (roughly 15 degrees per hour). If the heading is known, then the velocity detected by the INS will determine the local Latitude.

True North

if the Latitude is known then by detecting the Earth’s rotation the system is able to align with True North.

This alignment of position and orientation is an iterative process, each relying on the progress of the other.


Of the many different designs of INS, each with different performance characteristics, there are two main categories used in aircraft: stabilised platform and strap-down.

Stabilised Platform INS

Three accelerometers are mounted on a platform and orientated (usually) north/south, east/west, and up/down. This platform is driven by gyros (two or three) to always maintain its alignment with these axes regardless of any movement of the aircraft. Analogue feeds can be taken directly from the accelerometers and gyros that are in direct proportion to acceleration, and changes in velocity and direction. Stabilised platforms have some disadvantages. The main design issue is “Gimbal Lock”. Gyros are usually mounted in three gimbals on bearings; this allows the aircraft (and gimbals) to rotate around the gyros without moving the platform. However, when two of the three gimbals align, and are effectively operating around the same axis, they can become locked together and be directly affected by movement around the remaining third axis. The solution is to complicate the system further by adding a fourth motorised gimbal, which is continuously driven to avoid alignment with the other three.

Because of the mechanical assembly and many moving parts, stabilised platform INS suffer from wear and tear, and friction causes the output to “drift” over time. Compensation for drift requires complex bearing assemblies and special lubricants. Maintenance of stabilised platform INS is complex, costly and time-consuming.

Strap-Down INS

Ring Laser Gyros and accelerometers are attached rigidly, or “strapped down”, to the frame of the aircraft i.e. as the aircraft moves, so does the INS platform - exactly. Three gyroscopes sense the rate of roll, pitch, and yaw; and three accelerometers detect accelerations along each aircraft axis. It integrates them to get the orientation, then mathematically calculates the acceleration the north/south, east/west and up/down axes, as a gimballed system would.

With few moving parts strap-down systems are easier to maintain and more reliable over time. They do require more accurate gyroscopes and greater computing power. However, the advantages of reduced cost, size, weight and reliability make these systems the preferred choice of INS for aircraft.

Accuracy and Limitations

INS accuracy will drift with time; this can occur for many different reasons. Older stabilised platform INS (1970’s and early 1980’s) may have position errors of 2 nautical miles (nm) per hour. Modern strap-down LRG INS tend to have error rates of 0.6 nm/hr.

An INS is, generally, not the sole means of navigation on commercial airliners. INS position errors can be reduced by frequent updates of position from GPS and ground-based navigation aids. GPS tends to be given primacy over INS, by the FMS, in determining the aircraft position. However, when GPS signals are lost, or become unreliable, then INS can retain useful FMS position accuracy until the GPS reception improves. Also, it is common nowadays for GPS data to be used by INS for correction of some of the errors described below.


is mostly associated with imperfections. In stabilised platform INS, these imperfections exist within gyro bearings and mass imbalances. In strap-down RLG devices, the cause is derived almost entirely from imperfections in the mirrors and their coatings. Small imperfections in accelerometers cause small errors in acceleration sensing and these are compounded by the integration of these signals, first to velocity and then to distance. This is known as Integration Drift.

Transport Wander

a gyro will attempt to maintain it’s original alignment with the Earth. When the aircraft moves across the globe, the local horizontal reference changes and this will create an angle with a gyro. This error is known as Transport Wander and also Apparent Drift; it is usually corrected by signals derived from the INS Latitude and Longitude.

Schuler Loop

as the inertial platform accelerates in any one plane, it not only detects this linear acceleration, but (like somotogravic illusions in the inner ear) the platform will tilt and detect a component of Gravity, which increases the sensed linear acceleration. The opposite occurs during decelerations. Also the INS will try to compensate for this additional tilt in the same way as it compensates for Transport Wander. The system never catches up with itself and is always “chasing” itself. This loop is known as the Schuler Loop, and anywhere on the Earth the period of this loop is 84.4 minutes. What this means is that all things being equal and in a perfect INS with no drift, the position error would be greatest after 42.2 mins, and zero after 84.4 mins. In reality, the position error grows with each cycle of the Schuler Loop.

Earth Rotation

in the same way that a stationary INS detects the Earth’s Rotation to determine its Latitude and Orientation, it also detects this rotation when moving. The final movement detected will be a combination of both the aircraft’s movement and the Earth’s rotation. These errors are small and can be compensated for.


as an aircraft travels from A to B around the globe it actually will follow a curved path (or a series of shorter curved paths). An INS will detect this as turning (a bit like a pendulum swinging out in a turn) and an error can be introduced.

Inputs to an INS

Although an INS is essentially a “self-contained” navigation system, it does perform better when provided with some data inputs. These may include:

  • Initial Position: as the “shut-down” position will undoubtedly carry some error.
  • GPS Position: bounding the INS position error.
  • Barometric Altitude: helps to stabilise the vertical velocity and inertial altitude outputs.
  • True Air Speed (TAS): allows for the calculation of Wind Speed and Direction.

Inertial Outputs

As well as raw acceleration data in all planes and axes, typical Inertial outputs fed to other aircraft avionics and systems include, but are not limited to:

  • Pitch and Roll
  • True and Magnetic Heading
  • True Air Speed (TAS) and Ground Speed (G/S)
  • Latitude, Longitude and (inertial) Altitude
  • Wind Speed and Direction, and Drift Angle
  • Pitch, Roll and Yaw
  • Vertical Speed and Rate

Systems and Avionics utilising Inertial Outputs

Further Reading

It is recommended that, to understand the design and function of a specific INS, that the manufacturer's operating manual and flight manual are referred to; and, that for operational use the Air Operator's Operations Manual is referred to.


SKYbrary Partners:

Safety knowledge contributed by: