Sensors of the same type don’t necessarily measure the same “things”. Our latest blog post compares single- and multi-axis sensors to help you decide the right configuration for your application.
Tuesday, February 05, 2019
When evaluating different motion sensors, consider the number of axes your device needs to measure for your application. Then, consider the possibilities that could be unlocked by measuring additional–or different–axes. Choosing the right sensor capability early in your development process will lead to the best cost/performance fit, quick time to market and low development and product costs.
The simplest MEMS (Microelectromechanical System) motion sensing device is a single-axis sensor, such as a gyroscope or accelerometer. Accelerometers measure vibration or change in velocity while gyros measure the change in a device’s orientation (pitch, yaw and roll) by measuring angular velocity.
But, as the name implies, a single-axis motion sensor can measure motion on only one axis (x, y or z). So, a single-axis sensor can measure acceleration or rotation on a single plane, but it can’t determine device orientation or direction relative to any starting or absolute point. This can be useful if your application is generally used in one or two dimensions, where extra information is irrelevant. By using a single axis sensor, you can simplify processing and reduce costs. Just be sure that you don’t need additional context to allow your application to perform at its best. Otherwise, to add that type of information, you need to use several single-axis sensors in combination, or a multi-axis sensor.
Other types of (non-motion) single-axis sensors include thermometers and barometers.
- 2D surfaces (ground-based applications, such as robotic vacuum cleaners)
- Combined with other sensors to create an Inertial Measurement Unit (IMU) or add context (thermometer-how hot/cold was it, barometer-how high up)
- Used on their own to measure a specialized behavior (looking for a bump/rotation in a particular dimension)
A 3-axis sensor can measure angular motion on three axes (x, y and z). Most accelerometers, gyroscopes, and magnetometers have this measuring capability. This is not without good reason. If you want to track a particular input in three dimensions, you’ll want all 3 axes.
- Accelerometer: navigation & transportation, consumer electronics, industrial applications
- Magnetometer: mining & mineral exploration, oil & gas applications, smartphones, military applications, magnetic surveys
- Gyroscope: smart tech, industrial applications, GPS and navigation, consumer electronics, AR/VR applications
- Any 3-axis device: Combined into IMUs
Typical IMUs are 6-axis, composed of a 3-axis gyro with a 3-axis accelerometer. This configuration delivers relative heading, meaning change in heading as compared to previous heading, but not absolute heading. It also delivers relative orientation or tilt information (pitch and roll), as compared to previous orientation. Once you get into IMUs with multiple sensor inputs, sensor fusion helps to make sense of the raw sensor output.
One downfall of 6-axis IMUs is that without a reference point, they can be prone to a gradual build up of small errors over time, which lead to a drift in heading. This can be seen in the real world when robots slowly veer off their intended course.
- Navigation (dead reckoning, etc.)
- Consumer electronics motion tracking (AR/VR, smartphones, domestic robots)
- Industrial applications
- Aerospace & aviation applications
A 9-axis IMU adds information from a 3-axis magnetometer to the gyroscope and accelerometer. The magnetometer measures magnetic fields, delivering a fixed point of reference (Earth’s magnetic field). This data can be fused with the gyroscope and accelerometer data to deliver absolute heading: Not only how many degrees heading have changed, but its relation to magnetic north. In the same way, 9-axis devices also measure attitude (yaw, pitch, roll), or absolute orientation against a frame of reference.
Advanced sensor fusion of additional axes also enables compensations for drifts over long time periods, delivering more robust dynamic heading than a 6-axis alternative. Many common applications of a 6-axis IMU could see expanded capabilities with a switch to a 9-axis IMU. Drones, for example, can orient themselves to travel in a certain cardinal direction to follow the sun over the hills, or search an area more effectively.
- AR and VR gaming
- Robotics and autonomous vehicles
- Attitude alignment
Some sensors may have other numbers of axes–four, seven, ten, etc. These correlate to additional types of single-axis sensors. These single-axis sensors often collect data other than position or motion, including liquid levels or such environmental factors as temperature, humidity or air pressure. This environmental information adds important context to the motion information, providing data that can help a motion sensor perform optimally in varying conditions in the real world.
- Automotive applications
- HVAC applications
- Medical industry applications
- Outdoor sports & activities such as hiking
Each type of sensor has its uses. But, today’s sophisticated devices typically benefit from more data rather than less. Multi-axis IMUs can get the job done, but understanding their capabilities relative to your needs will enable you to choose the right one for your application.
If you want to talk about your choices, contact us. We work with development teams to determine the right sensor and sensor fusion configuration for their application.
You might also like
More from Sensor fusion
Navigation is a critical component of any robotic application. This post dives into the two of the most common tools …