Choosing the right Inertial Measurement Unit (IMU) for an aircraft, drone, missile, or spacecraft requires engineers to match the sensor’s capabilities to the vehicle’s mission and physical environment. Aerospace IMU selection helps determine the achievable accuracy of a navigation or stabilisation system, while also influencing size, power, qualification effort and supply risk. A sensor that performs well on a datasheet can behave differently under aircraft vibration, temperature changes and shock. Replacing it late in a programme may require further interface, calibration, software or test work, depending on how the new device differs from the original.
This post follows our Aerospace Vibration Electronics: Designing for Vibration and Shock Qualification post, which covered resonance, mounting and qualification testing. Here, the focus is on comparing IMU options and what to ask suppliers, with technical guidance for engineers and procurement questions for programme managers.
What an IMU does in an aerospace system (IMU, INS, AHRS and ADIRU)
An aerospace IMU selection system combines accelerometers (which measure specific force) and gyroscopes (which measure angular rate). Its measurements can be processed to estimate changes in velocity and orientation. The UK dual-use export list uses the term ‘inertial measurement equipment or systems’ for equipment of this kind, and its definition includes IMUs, inertial navigation systems, inertial reference systems and inertial reference units, as the published legislation shows. An inertial navigation system (INS) estimates navigation states such as velocity and position; an attitude and heading reference system (AHRS) focuses on attitude and heading, while an air data inertial reference unit (ADIRU) combines inertial-reference and air-data functions.
The performance of the finished system depends on processing as well as the sensors. SBG Systems, for example, advises against using raw IMU specifications to estimate the performance of its products when they run as an AHRS or INS, because filtering and estimation can affect the resulting outputs, as its raw IMU specification page explains.
Performance grades and sensor technologies (tactical, navigation and inertial grade)
Grade is shorthand for a broad performance class, not a universally standardised category. An industry overview of MEMS gyroscopes notes that accelerometers across a range of performance levels commonly use MEMS designs, while gyroscopes span a wider range of technologies. Ring laser gyroscopes (RLGs) and fibre-optic gyroscopes (FOGs) use the Sagnac effect. Their cost and performance vary by design and specification, so broad cost comparisons with MEMS devices should be treated cautiously.
|
Grade |
Gyroscope bias stability (indicative) |
Technologies and notes |
|
Tactical (indicative) |
Example: 6.25 °/h bias instability |
MEMS or FOG. Analog Devices lists a MEMS IMU at 6.25 °/h bias instability and a low-end FOG at 3 °/h; compare only where test methods and metrics align. |
|
Navigation (indicative) |
Example figures vary by supplier |
FOG, RLG or some high-performance MEMS designs. Accelerometer bias-stability figures of a few tens of micro-g are reported for some products; verify the metric and test conditions. |
|
High-grade inertial |
Below 0.001 °/h in some designs |
RLG and high-performance FOG designs are among the technologies used in demanding inertial systems. Technology alone does not define performance or application suitability. |
The figures in the table are illustrative examples drawn from Analogue Devices, SBG Systems and IEEE TechNav. Suppliers use grade labels and measurement methods differently, so compare the underlying specifications and test conditions rather than treating the thresholds as universal definitions.
Some higher-performance MEMS designs publish figures associated with tactical-grade performance and, in particular cases, more demanding classes. That does not establish navigation-system performance on its own: the result also depends on calibration, processing, aiding, integration and the conditions in which the device is used. Production maturity and supply arrangements should be assessed for the specific device.
|
🛠️ DESIGN CONSIDERATION Ask which test method, averaging time and temperature conditions sit behind each quoted figure. Bias stability, bias instability and in-run stability can be defined and measured differently between suppliers, so a comparison needs figures measured the same way. |
Parameters to specify in aerospace IMU selection (gyroscope, accelerometer and system level)
A specification can group IMU parameters into four sets:
- Gyroscope stability and noise: Bias stability characterises the stability of the sensor’s bias over time, while angle random walk characterises the effect of rate noise on integrated angle, as an IEEE overview of gyroscopes explains.
- Accelerometer performance: Bias stability and velocity random walk describe different aspects of accelerometer error. SBG Systems gives examples of navigation-grade accelerometers with bias-stability figures of a few tens of micro-g; check the exact metric, test method and conditions before using such figures as a benchmark.
- Scale factor and cross-axis sensitivity: These error terms affect how accurately the sensor reports input along each axis. According to the SBG glossary, navigation-grade IMUs may be calibrated individually on multi-axis motion tables across the operating temperature range; confirm the procedure for the specific device.
- Range and bandwidth: The measurement range and bandwidth need to cover the platform’s dynamics, and the output rate and latency need to suit the control loop that uses the data.
Datasheet figures describe the sensor under defined test conditions. Installed performance can differ, particularly where the mounting structure, vibration spectrum or operating temperature differs from those conditions.
How vibration, shock and temperature change IMU performance (vibration rectification and installation effects)
Vibration rectification error (VRE) describes a bias produced when vibration interacts with a sensor’s response. A review of the effect in MEMS accelerometers notes that the resulting offset can contribute to navigation error over time. In gyroscopes, vibration can also produce an angular-rate bias; the magnitude and behaviour depend on the device and test conditions, as Inside Unmanned Systems reported.
Suppliers may express gyroscope vibration sensitivity or rectification in units such as degrees per hour per g squared, but definitions and test methods need to be checked. One published MEMS IMU specification, for example, quotes gyroscope vibration rectification error below 1 °/h/g² under a stated test condition, according to SBG Systems’ documentation. Treat such figures as comparable only when the vibration level, frequency range, axis and measurement method are sufficiently similar.
Installation adds further effects. The mounting structure and any isolators determine the vibration transmitted to the sensor. Where invoked by the applicable equipment specification or certification basis, RTCA DO-160 provides environmental test procedures for airborne equipment; Section 8 addresses vibration. The applicable category and test conditions depend on the installation and qualification requirements. Passing an environmental test does not by itself establish navigation accuracy under every installed condition. Aerospace Vibration Electronics: Designing for Vibration and Shock Qualification covers resonance and mounting design in more detail.
|
⚠️ CRITICAL ALERT Datasheet vibration figures may be measured at levels and frequencies that differ from the installation. Ask for rectification and bias-shift data over a spectrum comparable to the installed environment, and consider testing a shortlisted sensor on a representative mount before the selection is frozen. |
Qualification and design assurance evidence (what an IMU supplier may provide)
Qualification expectations depend on the system the IMU sits in. Environmental qualification may be specified through a TSO, equipment specification, customer requirement or certification basis; RTCA DO-160 is one common reference. Where the IMU forms part of a system subject to DO-254 hardware or DO-178C software design-assurance objectives, the supplier evidence available can affect the work required from the integrator. These activities should be assessed separately: environmental qualification does not replace hardware or software design assurance.
A DTIC-hosted study of COTS components and DO-254 discusses the challenge of using previously developed hardware within a design-assurance process. Existing supplier design data may support that process, but it does not automatically establish acceptability or discharge the integrator’s responsibilities. Some vendors of COTS avionics modules supply supporting documentation packages, as Military Embedded Systems describes for processing and graphics modules. Whether comparable evidence exists for a particular inertial sensor needs to be confirmed with its supplier.
Export control and supply (what can change the shortlist)
Inertial measurement equipment is addressed in Category 7 (Navigation and Avionics) of the UK Strategic Export Control List. Entry 7A003 contains performance-based criteria and technical notes. Its note states that 7A003 does not control inertial measurement equipment or systems certified for use on civil aircraft by civil aviation authorities of one or more EU Member States or Wassenaar Arrangement participating states, subject to the applicable wording and jurisdictional provisions in the current list. This should not be treated as a blanket exemption based solely on intended civil use. Classification and any licensing requirement depend on the specific item and circumstances, so check the current list and obtain appropriate export-compliance advice before committing to a sensor version published on legislation.gov.uk
Export controls can affect lead times and the ability to supply particular destinations, but the classification of a sensor does not automatically determine the status of every finished product that incorporates it. Assess the finished system and any applicable controls on incorporated equipment, software or technology. Items of non-UK origin may also be subject to controls in other jurisdictions.
Case Study: Qantas Flight 72 and the Limits of Sensor Selection
On 7 October 2008, an Airbus A330-303 operating as Qantas flight 72 was cruising when one of its three air data inertial reference units (ADIRUs) began producing spikes of incorrect data. The flight-control primary computers commanded a pitch-down. The Australian Transport Safety Bureau concluded in investigation AO-2008-070 that the upset resulted from a design limitation in the flight-control software combined with a failure mode affecting one of the three units. Australian Aviation reported that 119 people were injured.
Flight Safety Australia’s account of the investigation adds that the spikes were associated with intermittent problems in the unit’s central processor. The built-in test equipment was not designed to detect this particular failure mode in the air-data parameters, so the affected output was not flagged as invalid. The flight-control computers’ algorithm for managing incorrect data could not handle spikes spaced 1.2 seconds apart. The aircraft manufacturer subsequently redesigned the angle-of-attack algorithm.
The event was not a straightforward case of an IMU’s inertial measurement accuracy degrading under vibration. The investigation identified a data-spike failure mode in an ADIRU and a limitation in the flight-control computers’ handling of particular angle-of-attack data patterns; it did not establish what triggered the ADIRU fault. The outcome illustrates how fault detection, validity signalling and host-system handling can interact. A comparison based only on bias and noise figures would not capture those system-level behaviours.
What aerospace IMU selection can cost a programme (and where a structured evaluation helps)
A sensor change after integration can reopen interface, calibration, software and test work. The extent depends on differences in output format, mounting, noise behaviour and available qualification evidence. Late discovery of an export restriction, a long lead time or vibration sensitivity not covered by the datasheet can create additional programme risk.
Before freezing the selection, compare candidate IMUs against the system error budget and installed environment. Request supplier data on vibration response, integrity monitoring and qualification, review export classification and supply arrangements, and consider bench or vibration testing on representative mounts where the available evidence leaves important questions unanswered.
Questions to ask an IMU supplier (before you commit)
Programme managers and buyers can assess an IMU supplier’s readiness with four questions:
- Installed performance: What data supports the quoted bias, noise and vibration-rectification figures across the vibration, shock and temperature range of this installation, and how were they measured?
- Integrity and failure modes: What built-in test coverage, validity flags and failure-mode information are provided, and how should the host system monitor the output?
- Qualification and design evidence: Which environmental test reports and hardware or software design-assurance data are available, and which gaps would remain for the integrator to address?
- Supply and export: What are the item’s export classification, lead times and end-of-life arrangements, and is a second source or form-fit-function alternative available?
Final Thoughts
The performance class and technology help shape theaerospace IMU selection error budget, but installed vibration and temperature behaviour can differ from the published figures. Qualification evidence, integrity monitoring, export classification and supply arrangements also need to be assessed for the intended application.
Qantas flight 72 illustrates how a unit’s failure behaviour and the host system’s response can affect the outcome. For IMU selection, supplier evidence on installed performance, fault indication and qualification gives the programme a more reliable basis for comparison. The next post in this series, Test Rigs for Aerospace Validation, looks at how the environmental performance of airborne equipment is assessed on the test bench.
|
WORK WITH IGNITEC Ignitec’s engineers support airborne electronics programmes from early design through to qualification planning. If you are comparing IMU options, we can review your error budget and installation environment, and identify vibration and integration risks before you commit to a sensor and test plan. |
Contact the team to discuss your programme and IMU selection strategy.


