Sensor Identifying EMI Sources Improves IoT Safety

March 16, 2017 by Jeff Shepard

Kanazawa University and Hitachi, Ltd. announced the joint-development of a compact sensor capable of identifying the source of electromagnetic interference that causes malfunction in component for autonomous operation. The realization of the compact sensor will not only allow multiple sensors to be embedded in vehicles such as automobiles and trains, but also by configuring it with several antennas, enable the direction of this low-frequency electromagnetic interference emitted by aging components to be precisely detected, and thus allow the accurate identification of the electromagnetic interference source.

By using this sensor, the risk of malfunction in autonomous operation component or transport systems can be avoided by continuously monitoring electromagnetic interference, and preventing the loss of sensor data related to cruise control. Kanazawa University and Hitachi will continue with the application of sensing technology detecting electromagnetic interference, to contribute to improving safety in societal infrastructure.

In the recent years with the progress of IoT, initiatives are rapidly growing to connect components embedded in automobiles or trains, or autonomous operation. To ensure the long term reliability and safety of aging component and systems, it is necessary not only to increase the reliability of the components themselves but also to obtain normal sensor data. Low-frequency electromagnetic interference that interferes with surrounding signals and sensor data, however, leaks from aging or integrated electronic components that may result in equipment or system malfunction.

Currently, there are two technologies to identify the source of low-frequency electromagnetic interference: (a) the application of high-frequency direction-of-arrival (DOA) estimation used in wireless communication, and (b) the application of DOA estimation using Poynting vector. High-frequency DOA estimation, however, requires either larger sensors or the installment of sensors over a wide area, etc., therefore making practical application in autonomous operation component difficult.

On the other hand, DOA estimation using Poynting vector measures either the electric or magnetic field, and obtains the other value by conversion to determine the probable direction of the source, producing an error of approximately 45 degrees preventing the accurate identification of the electromagnetic interference sources.

To precisely identify the source of low-frequency electromagnetic interference, Kanazawa University and Hitachi developed a compact sensor applying the Poynting vector method that can simultaneously measure both the electric field as well as the magnetic field in three dimensions from the same location. Features of this sensor are as follows:

1. Two loop antennas arranged in a pair structure to enable simultaneous measurement of electric and magnetic fields - Two loop antennas, conventionally only used in magnetic field measurement, are arranged as a pair structure on the same plane to enable simultaneous measurement of the electric field occurring between antenna.

2. Three pairs of loop antennas arranged to cross each other achieves sensor's high accuracy and small-size - Three paired loop antennas are arranged so that the circuit for detecting electromagnetic interference set in the center can obtain all the data at once. This allows the data on the electric and magnetic fields to be simultaneously detected in three dimensions. As the loop antennas are aligned perpendicularly, measurement accuracy is improved without interference from each other. In addition, as the detection circuit is integrated in the center of the pair structured loop antenna, a compact size of 110 cubic mm was realized.

Using the sensor developed to identify the direction of electromagnetic interference of 200kHz to 10MHz, detection with an error of less than 2 degrees of the actual direction was confirmed. This converts to an error of less than 10cm when the distance between the sensor and the noise sources is 3m. By using three sensors, it is possible to specify the source of electromagnetic interference as the point where the three detected directions cross, thus allowing component malfunctions to be prevented.

Kanazawa University and Hitachi, by applying this sensor to the establishment of error prevention systems for autonomous operation component, will contribute to enhancing the safety of societal infrastructure.

A part of this achievement will be presented at the IEICE General Conference to be held from 22nd to 25th March 2017 at Meijo University, Nagoya, Japan.