Do you know what an IMU is? Do you know what it stands for? And do you know how it actually works?
If you work in automotive testing, autonomous navigation, mapping, or even in aerospace and defence, chances are you’ve come across the term “IMU” in your work. In this blog, we’re explaining what they are, why they are important, and how they work.
IMU stands for inertial measurement unit
There you go. An inertial measurement unit can either exist as a standalone component of a navigation engine, or integrated into an inertial navigation system, or INS. That’s what we do here at OxTS – our INS devices feature an IMU fused with a GNSS receiver, with both working in harmony to give you accurate position and orientation data.
What do you use an IMU for?
Inertial measurement units help you measure the dynamics or motion of an object.
IMUs are vital for vehicle testing
Automotive manufacturers need to test their vehicles and the systems in them to rigorous standards. An IMU (as part of an INS) gives manufacturers precise information about their vehicle at every stage of a test. For instance, during an Autonomous Emergency Braking test, the IMU can tell you how fast the vehicle was going when the system deployed, exactly which way it was pointing, the rate of deceleration, and more.
Open road testing is also becoming more and more important – in these environments, an IMU helps ensure that you can accurately record the position of your vehicle at all times.
Learn more:
IMUs enable autonomous navigation
Whether we’re talking robotaxis, automated fruit pickers, or UAVs, IMUs are fundamental to autonomous navigation. While all these vehicles will use GNSS signal to achieve position information, that signal won’t always be available.
Sometimes a vehicle will go through a tunnel, into a forest, or even into an area with GNSS jamming in place if we’re talking about defence applications. In those environments, the IMU helps keep the vehicle on track through dead reckoning. The INS takes the last known position from the GNSS data, and the data from the IMU about how fast it’s moving and in what direction, and uses that to estimate the current position.
The image here, shows an autonomous development robot we built internally. The robot uses an AV200 INS (IMU and GNSS) which enables the robot navigate accurately in both indoor and outdoor environments. The robot uses GNSS signal to calculate its position in outdoor environments and when GNSS signal isn’t available, the robot can continue to navigate using the IMU and data from other onboard sensors such as camera and LiDAR using the OxTS GAD Interface.
Learn more:
IMUs keep position data accurate for georeferencing activities
If you’re conducting a survey over a large area, whether by land or by drone, accurate position data is vital. An INS is the centre of many surveying payloads, enabling the user to calculate their position at every point during the survey so that the data can be georeferenced post-mission. Just like with navigation, the IMU helps keep the INS data accurate by providing a high-frequency stream of data to help calculate the payload’s position and its heading, and orientation.
To see how an IMU contributes, consider these two images. The yellow KML trail is the position of a car driving around London based purely on the GNSS data:
You can see lots of erratic, uneven lines and the trail running through buildings – in other words, inaccurate data. Now, here’s the same route but with data from the OxTS IMU added in:
As you can see, the data from the IMU vastly improves the accuracy of the position data the INS generates.
It’s not perfect – London is one of the most challenging environments to get an accurate position in, due to the volume of urban canyons and tunnels – but the good news is that OxTS can do even better, as we’ll show you below.
Learn more:
Intelligence, surveying and reconnaissance (ISR)
So, now we know what IMU means, and what an IMU can do. But how does it actually work?
What is an inertial measurement unit?
The simple answer is that an IMU measures changes in linear acceleration and angular velocity. When it’s integrated into an INS, that data is used to estimate the speed, heading, and orientation of the IMU (and anything that’s attached to it).
You could think of it this way: you know the feeling of being pushed back into your seat when your car accelerates or pushed towards one side as it goes around a corner? An IMU experiences similar pushing forces and can use that data to calculate how it is moving.
An IMU measures those things using a collection of accelerometers and gyroscopes. But not all IMUs are made the same as we’re about to discuss.
Different IMU technologies
The most popular IMU technology is MEMS, which stands for micro-electro-mechanical systems. That means that the entire IMU is small enough to fit onto a circuit board, which makes MEMS technology ideal for applications where size is a factor such as in smartphones or on UAVs. They can be made of quartz, but most are made of silicon.
There are also FOG and RLG IMUs (those stand for fibre optic gyroscope and ring laser gyroscope respectively). These IMUs, as the name implies, use a different type of gyroscope to measure angular velocity – the accelerometers may well be the same designs that you’d find in a MEMS IMU.
RLG and FOG IMUs are bigger and more expensive then MEMS IMUs, and have long enjoyed a reputation for being more accurate and stable (meaning their accuracy takes longer to degrade) than MEMS IMUs. However, MEMS technology is improving all the time – and at OxTS, we’ve spent years developing precise calibration techniques and advanced data processing algorithms to squeeze the best performance from our MEMS IMU technology. We’ve also focused on using a variety of additional sensors to further improve accuracy, giving our users FOG-level performance using MEMS technology (and at MEMS prices!)
To show you what we mean – remember that image of the position data for a journey around London? The IMU makes the data much more accurate than GNSS data by itself, but there are still errors – you can see lines running through buildings, for instance.
One of the innovations we’ve been working on at OxTS is LiDAR inertial odometry, or OxTS LIO for short. It uses velocity data from a LiDAR scanner to calculate orientation and velocity data which can be factored into the calculations happening inside the INS. Adding this data in on top of the IMU produces position data that looks like this:
Much more accurate – and done without an expensive FOG or RLG IMU.
We’ve only scratched the surface
This blog is aimed at those who are new to the field of inertial navigation – so it goes without saying, there’s a lot more to talk about. How does the spin of the earth affect my IMU? What do you do if your IMU and GNSS measure different distances? How does all this connect to PNT? Your journey into the intricacies of inertial navigation are only just beginning…
You can check out the blogs linked above to learn more about inertial navigation – or talk to our team about your project and your questions. We love to talk about this stuff! Just contact us using the form below and we’ll be in touch!
