Talk:Inertial measurement unit

This page
This page is pretty much redundant with Inertial guidance system, which is more detailed


 * Why? Inertial Measurement Unit is a part of the Inertial guidance system. I believe it is the hardware that actually performs the measuring i.e. the gyroscopes and the accelerometers. However, that is not clearly explained in this article. I don't know enough about it but maybe someone else feels up to the task?


 * from a user without a wikipedia log:
 * To my experience, the term IMU is widely used to refer to a box, where: 3 accelerometers are placed such that their measuring axis are orthogonal to each other, measuring the so called specific forces (inertial acceleration - gravity); 3 gyros are placed such that their measuring axis are orthogonal to each other, measuring the rotation rates; one or more temperature sensors are included, maybe already incorporated in each acc or gyro, or as an additional sensor which data will be used to calibrate the raw data from the other sensors. To achieve superior accuracy, the box may be designed such that the temperature is controlled and kept constant, and the walls of the box are made of materials that minimize electromagnetic interference. If the output signals are analog, further considerations must be taken about the cables and the A/D card. If the output data is already in digital format, time delays become the major concern that should be carefully modelled.
 * The data provided by an IMU box is all you need to perform dead-reckoning. It's not said on the article, but perhaps the first use of such box was in a ship, and still today almost every ship has one. Satellites have one also. Almost anything that must somehow use some electronics to know its acceleration, velocity and/or attitude, have an IMU. Nowadays you can buy an out-of-the-shelf MEMS based IMU for something like 2000 eur, and that's a good one already for many purposes.
 * If we work a bit the kinematic equations and let it run on a Kalman filter, the data from an IMU can be transformed into an ARS system (Attitude Reference System). This means roll and pitch. The dead-reckon integration of the rate gyros will cause drifting errors, but the observation of the gravity vector by the accs serves as an external observation of the local vertical, which corrects for these drifting errors. In dynamic environments like in a jet fighter, the gravity will be masked by the aircraft's body accelerations, thus normally in these cases the IMU is coupled together with the GPS or other sensors. This brings us a bit closer to a Navigation and Guidance System and a bit away from the IMU. Maybe this is why a lot of engineers don't really make a distintion.


 * well, perhaps some good english writer and more baldy than me, will make this official... feel free


 * - Reply on your text: An INS and an IMU are different products. IMU's are also used for human motion tracking (for rehabilitation, animation and Virtual Reality for example). An INS is used for machine motion and IMU's are indeed part of INS's, but the use of IMU's is not limited to this appliction.


 * An INS may have several IMU's, for redundancy. Also, an attitude estimation system may also use an IMU, but is not an INS. —Preceding unsigned comment added by Alan.A.Mick (talk • contribs) 15:37, 18 November 2010 (UTC)


 * I revamped the article to address all these issues. Also, to those who posted to the former "external links" section, I've removed that section because it only had corporate websites listed and was becoming a kind of free ad space. Generally, we should probably limit external links to scientific information resources. Equazcion (Talk • Contribs) 20:26, 4 September 2007 (UTC)


 * The above are correct. An IMU refers to the sensors (typically accels and gyros, hence Inertial Measurement Unit).  An INS refers to compilation of many sensors (One or more IMUs, GPS receiver, etc.) and the accompanying software to generate estimated acceleration, velocity, and position (hence Inertial Navigation System).  Part of the confusion is the disparity in age, accuracy, and cost.  Old heads and military/space/aviation folks think of mechanically stabilized gyros or ring laser gyros, while young bucks think of MEMs.  The textbook/authoritative sources mostly reference the older technology. Wiki's, DIY tutorials, and modern journal articles (especially for robotics or wearables) mostly reference MEMs. While the underlying math is technically the same, the two worlds are so vastly different that its almost a disservice to speak of them in the same terms.  Add to that the fact that MEMs are rapidly approaching navigation grade, and it only further muddies the waters--Keithicus (talk) 16:40, 26 June 2018 (UTC)

Accuracy of measurements
How much accuracy can you expect from an IMU? --Singularitarian (talk) 18:18, 23 May 2008 (UTC)

Like a lot of other engineering components, there are a wide variety of configurations and various degrees of accuracy and percision at various price points. There realy is no general answer to that question. —Preceding unsigned comment added by Alan.A.Mick (talk • contribs) 15:33, 18 November 2010 (UTC)

Another kind of IMU
There's another kind of IMU called "gyro-free". Indeed, 6 or more accelerometers in a "good" configuration could be used to measure linear acceleration and angular acceleration. All this in "Feasibility of A Gyroscope free Inertial Navigation System for Tracking Rigid Body Motion" Tan, Mostov Varaiya (200), Institute of Transportation Studies, University of California, Berkley.

Also, I consider that an INS is a system that generally contains and IMU but it contains another subsystems like a GPS, a computer, a memory bank for map data and so on. The IMU could be called the sensor part of the INS. Crodrigue1 (talk) 17:40, 21 May 2009 (UTC)

TIMU (Timing & IMU) sensors
DARPA's Microsystems Technology Office( MTO) department is working on a Micro-PNT ("Micro-Technology for Positioning, Navigation and Timing") program to design "TIMU" ("Timing & Inertial Measurement Unit") chips that does absolute position tracking on a single chip without GPS aided navigation.

Micro-PNT adds integrates a highly-accurate master timing clock integrated into a IMU (Inertial Measurement Unit) chip, making it a "TIMU" ("Timing & Inertial Measurement Unit") chip. So these TIMU chips for Micro-PNT have integrated 3-axis gyroscope, 3-axis accelerometer, and 3-axis magnetometer, and together with the integrated highly-accurate master timing clock it simultaneous measure the motion tracked and combines that with timing from the synchronized clock, and with sensor fusion it makes a single chip that does absolute position tracking, all without external transmitters/transceivers. 164.4.17.36 (talk) 12:04, 26 June 2013 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Inertial measurement unit. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20130630051230/http://www.darpa.mil/NewsEvents/Releases/2012/09/20.aspx to http://www.darpa.mil/NewsEvents/Releases/2012/09/20.aspx

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 17:35, 13 November 2017 (UTC)

Create applications section
Much of the technical discussion incorporates discussion of specific fields of study where IMUs are applied. I think it'd be worthwhile to break these out into an applications section e.g. driverless cars, wearables, robotics, consumer electronics, etc.--Keithicus (talk) 16:44, 26 June 2018 (UTC)

Sensor Errors
The last sentence the section is "Calibration will typically improve sensors raw performance by at least two decades." Decades makes no sense for a sensor. Perhaps it should read decibels. The first reference from google scholar for 'inertial measurement unit' finds "Calibration of a MEMS inertial measurement unit" at https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.65.5896&rep=rep1&type=pdf From the conclusions of the paper: "Simulation results shows that the mean square error of the parameter estimates of the senor model increases with up to 8 decibel, when utilizing the proposed method." Perhaps someone that knows what they are talking about can look at this. I'm not that person. — Preceding unsigned comment added by 2603:8001:7C3F:A436:E04E:879E:6202:EF08 (talk) 17:36, 7 March 2021 (UTC)