Inertial Sensors for Navigation, Guidance and Control of Vehicles
General Project Information
Overview
Accurate and cheap vehicle navigation systems are essential for many Intelligent
Transportation Systems (ITS) applications. Current navigation systems used in aircraft cost several times
the price of an average car. Moreover, these expensive systems do not provide the accuracy needed to
navigate a vehicle within a narrow lane at speeds of 90 miles/hour. The PATH navigation system must
provide measurements of absolute and relative position, velocity, and acceleration to automated vehicles.
This system will eventually use inertial measurements, map matching, satellite positioning, and magnetic markers.
Collaboration Efforts
In this introduction I will concentrate on the description of the inertial measurement systems
currently under development in UC Berkeley. We believe our system will set a standard for low cost and high accuracy navigation units.
We are in a unique position to accomplish this goal because of our close alliance with experts developing state of the art sensor technology, namely silicon surface micromachining.
During the past several years, Professors Boser, Howe and Pisano of
BSAC
(Berkeley Sensor and Actuator Center) and their graduate students have been working to implement this technology in micromachined inertial sensors for PATH.
The BSAC group has pursued new approaches in integrated technology, including
innovative designs of sigma-delta accelerometers and vibrating gyroscopes. The Berkeley designs are fabricated at foundries of Analog Devices and at other high quality manufacturing facilities. Last year
Analog Devices
provided such sensors on a commercial basis. This opened up the opportunity to build viable systems for PATH.
System Design
Our effort focuses on developing vehicle navigation systems. The system level design
is divided into four main tasks. The first task is to develop detailed specifications for the sensors and to
design the general architecture of the IMU (inertial measurement unit), including its mounting in a
vehicle. The second task is the design of signal conditioning circuitry, analog filtering and A/D
conversion sub-systems. The third and most challenging task is to develop algorithms for sensor
calibration and data processing and to design software for an embedded DSP computer. The final task is to
develop methods for testing the performance of the inertial sensors and IMUs using system identification in
a testing facility. All four tasks have been pursued aggressively in 1995, and a first generation "proof
of concept" multisensor system has been built. It is shown on the inertial test stand in rate table photo.
Prototypes and Cost
Once developed, prototypes will be implemented on test vehicles and will be made available to other research groups. Currently we are focusing on the first generation six-sensor module that could replace single axis accelerometers in each PATH vehicle and will measure true acceleration for longitudinal control. The system will measure six degrees of freedom for each vehicle. The measurements can be used to improve the ride quality in a platoon. The IMU with accelerometers fixed in a cube configuration is shown in the cube photo 
The choice of integrated sensors has resulted in a ten-fold price reduction, with the total system component cost being under $100; a single accelerometer
currently costs over $1000. However, the first generation (ADXL05) integrated sensors suffer from numerous errors. To achieve the required performance, complex calibration and processing must be performed on the data provided by each sensor. Mathematical models and software algorithms have been developed to correct for sensor errors using a "generic" approach. The approach allows users to adjust parameters to accommodate sensors with different characteristics or a different number of sensors per system.
BSAC researchers are expected to provide a broad new array of linear multiple axis
accelerometers and angular rate gyroscopes with improved accuracy. Such sensors will greatly enhance
the current IMU performance and will allow us to configure systems for new compelling applications.
More Information
PATH Sensor Group /
Web Editor
