Highly Accurate Motion Sensing and Control for Industrial Applications

Many industrial applications either already use IMUs or would benefit from high-accuracy IMUs that cost less than the military-grade versions, which can easily cost more than $10,000. Examples of such positioning and motion control applications include:

  • Machine control in construction vehicles such as bulldozers, excavators, and graders for accurate positioning or leveling of the control arm, blade, bucket, or other end-effect tool.
  • Autopilot systems in agricultural machinery such as planting, seeding, and harvesting equipment to provide precision placement of material to maximize efficiency and yield.
  • Motion compensation or stabilization for satellite and camera platforms as part of a feedback control system to track or maintain the steady position of the pointing system.
  • Survey and mapping systems to allow accurate time-correlated positioning of collected scanned imaging data and to create accurate topographic or hydrographic surveys.
  • General vehicle motion analysis for the purpose of performance tuning, ride and handling analysis, and maintenance and crash safety.

IMU Features Overview
Although the definition of an IMU differs slightly among vendors, here we will define it as consisting of a triaxial angular rate sensor (gyroscope) and triaxial linear acceleration sensor (accelerometer). The IMU provides these three-dimensional sensor data to an external system to describe the IMU body's spatial behavior and motion without requiring an external reference.

Figure 1 shows the high-level block diagram of the Epson S4E5A0A0. Its key components and functional blocks include:

  • Precision temperature sensor used by the internal temperature compensation logic
  • Triaxial orthogonal gyroscope and accelerometers to sense the angular velocity and specific force
  • Analog front end and conditioning block, which filters and converts the sensor signal output for digital signal processing
  • Digital signal processing, which includes digital signal filtering, sensor data enhancement through calibration, data formatting (unit conversion), and diagnostic self-test
  • SPI or UART interface circuitry for connectivity to a variety of host platforms

 

Figure 1. Epson S4E5A0A0 IMU block diagram
Figure 1. Epson S4E5A0A0 IMU block diagram

 

QMEMS: The Key to Meeting Design Challenges
At the core of the Epson IMU is the company's QMEMS gyroscope technology. Vibration gyroscopes sense angular velocity from the Coriolis force applied to a vibrating element. For this reason, the accuracy with which the angular velocity is measured differs significantly depending on the material from which the sensing element is made as well as its structural differences.

Vibration gyroscope manufacturers use a variety of materials and structures to create compact, high-accuracy gyroscopes that provide low-noise characteristics, high shock resistance, a high scale factor (a measure of sensitivity), high stability, a low temperature-frequency coefficient, and a compact size.

Here is the basic operating principle of the Epson QMEMS gyro sensor.

  1. Under normal conditions, the drive arm vibrates in a specific direction (Figure 2).
  2. As the gyro is rotated, the Coriolis force acts on the drive arms, producing a vertical vibration (Figure 3).
  3. The stationary part of the sensing element bends in response to the vertical drive-arm vibration, inducing a sensing motion in the sensing arms (Figure 4).
  4. As the pair of sensing arms moves, they generate a potential difference, which is used to measure the angular velocity (Figure 5). The angular velocity is converted and output as an electrical signal.

 

Figure 2. Normal drive vibration
Figure 2. Normal drive vibration
  Figure 3. Body rotation
Figure 3. Body rotation

 

Figure 4. Produces vertical vibration
Figure 4. Produces vertical vibration
  Figure 5.  Induce and detect sensing arm vibration
Figure 5. Induce and detect sensing arm vibration

 

As illustrated in the figures, the QMEMS technology uses a precision quartz double-T structure that yields a Q factor (a measure of resonance efficiency) of approximately 30,000, where a higher Q factor is more desirable than a lower one. In comparison, an H-type tuning fork structure has a Q factor of about 20,000; a silicon tuning fork has a Q factor of about 10,000; and a typical silicon MEMS structure has a Q factor of 3,000 to 5,000.

The high-Q double-T sensing structure used in QMEMS devices results in excellent stability and performance for the Epson IMU. Additionally, the sensor's symmetry and isolated drive and sensing arms generates very little noise, and its center-of-mass anchor point provides vibration and shock isolation. The quartz sensing element, advanced analog circuitry, and the device's high Q factor enables its low noise characteristics and high signal-to-noise ratio (SNR). As a result, the gyroscope provides accurate angular rate sensing for rate changes from very small to very large, with minimal errors from short- or long-term drift.

Quartz material has long been used as a stable oscillator, and its resonance frequency is less sensitive to temperature effects in comparison to other materials such as silicon. The QMEMS gyro element offers stable sensing properties for scale factor (sensitivity) and bias (signal output under nonrotation).

IMU Considerations
Often, an IMU is a critical functional building block used in products such as an Attitude Heading and Reference System (AHRS) or Inertial Navigation System (INS). As shown in the typical INS block diagram in Figure 6, the IMU sensor data are processed by the system to obtain the system's position, velocity, and attitude.

 

Figure 6. Typical INS block diagram
Figure 6. Typical INS block diagram

 

In many cases, inertial sensors such as IMUs augment other forms of sensing, including Global Navigation Satellite Systems (GNSS) or optical-based technology for distance measurement, such as Light Detection and Ranging (LIDAR), because the IMU is immune to external interferences that are problematic for RF or line-of-sight technologies. For example, in the case of GNSS it is surprisingly common for blockage or multiple signal reflections from natural and man-made structures to cause reception interference. During these short-duration GNSS outages, the IMU can provide estimated positioning data by dead reckoning from the time of last GNSS fix. In precision aerial surveying applications, IMUs can provide precise orientation of LIDAR sensors that are rigidly mounted to an aerial vehicle for topographic mapping of the landscape below. Typically, multiple (aiding) sensor data are combined with IMU measurement data by a mathematical technique called a Kalman filter, which includes sensor error modeling and a recursive processing algorithm to produce optimal measurement results.

A significant challenge is inherent to all IMUs: They suffer from errors, including short-term bias error, scale-factor error, misalignment, noise (random walk), and bias instability. Extensive on-board calibration and temperature compensation integrated within the IMU mitigate the effects of these errors.

Typically, in industrial applications, more stringent performance tolerances require the use of fully calibrated inertial sensors to meet design targets in terms of error, such as bias, scale factor, axial alignment, noise, temperature effect, and repeatability from IMU to IMU. These performance metrics directly affect subsequent blocks in the signal chain and impact the overall performance of the system's filtering and control-loop algorithms.

Providing in-house sensor calibration requires a significant investment in equipment, development resources, and engineering time. Inertial test equipment such as precision multi-axis rate tables, air isolation platforms, vibration testers, and temperature chambers are price-prohibitive, due to their significant upfront cost, which is hundreds of thousands of dollars. In addition, significant development time and resources are associated with calibration algorithms, test methodology, and custom software suites, which can easily exceed tens of man-months. Due to the variability between MEMS inertial sensor characteristics and the required performance tolerances, having to include in-house calibration directly into the end-equipment manufacturing process poses a significant cost and process challenge (and may not be feasible).

Epson's design challenge was to develop an IMU that exceeded current target application performance requirements—in particular, those for accuracy and stability—while simultaneously shrinking the form factor to just under 1 in. × 1 in. × 0.5 in. and reducing the power consumption to <100 mW. With this combination of small size and low power consumption, the Epson IMU is designed to be a key enabler of applications and features not possible with current products on the market. In the case of the Epson S4E5A0A0, using the company's QMEMS quartz technology and optimized calibration IP was key to the IMU's performance and quality.

QMEMS and the IMU
In general, the performance of the gyroscope is the primary factor determining the overall performance of an IMU.

Epson's QMEMS-based IMU has the following characteristics:

  • 6 degrees of freedom (DoF)
    • Triaxial gyroscope with a ±300°/s measurement range
    • Triaxial accelerometer with a ±3 G measurement range
  • Low noise and high stability, as demonstrated by its angular random walk of 0.24°/√hr. and its gyro bias instability of 6°/hr.
  • Gyro sensitivity temperature coefficient of 25 ppm/°C
  • Gyro bias temperature coefficient of 0.005°/s/°C
  • Axis-to-axis alignment of 0.1°
  • Small size (external dimensions of 24 mm × 24 mm × 10 mm) and light weight
  • Power consumption of 30 mA at VDD = 3.3 V
  • Standard SPI/UART digital serial interface
  • Operating temperature range of –20°C to 70°C

Because it is the most accurate and compact IMU in its class—with precision and stability exceeded only by military-grade products—the S4E5A0A0 enables new motion sensing or control applications in a range of industrial applications and areas such as intelligent highway transportation systems; specialized tools or apparatus to assist in medical procedures that require precision manual alignment and orientation; or small, intelligent, autonomously controlled devices for commercial use in maintenance and security. It can also be used in many other applications where traditional optical-based IMUs are unsuitable due to either high cost, large size or weight, or high power consumption.

ABOUT THE AUTHOR
Bob Porooshani is General Manager of the Microelectronics Operations div. of Epson Electronics America Inc., San Jose, CA. He can be reached at 310-955-5319, [email protected].