The increasing replacement of mechanical parts by x-by-wire systems in automotive applications allows improving driver safety. These systems demand highly dependable sensors that ensure their functionality despite the harsh operating conditions. This means that the sensors should be capable of working continuously despite catastrophic faults and keeping the performance over time. An anisotropic magnetoresistance (AMR) sensor is a magnetic sensor commonly used for angle measurements in cars. It is affected by catastrophic faults and performance degradation due to undesired parameters included at the sensor outputs. Until now, physical redundancy is often used to handle catastrophic faults. For the performance, compensation factors for the undesired parameter, such as offset voltage, are estimated at the start of the sensor life. Although the undesired parameters drift due to aging effects, the sensor performance remains within the allowed tolerant band. However, this tolerant band will decrease in the future because the dependability requirements are continuously increasing. Therefore, it is necessary to consider strategies to guarantee the sensor performance over time. This paper proposes a system to improve the sensor dependability using analytical redundancy for catastrophic faults but also with self-x properties to maintain the sensor performance over time. Results indicate a dependability improvement in terms of reliability, with a reduction of 50% in the rate of uncovered failures. The safety requirement ASIL level D is satisfied, and with regard to maintainability, the sensor performance is maintained over time.
|Title of host publication||2017 International Test Conference in Asia (ITC-Asia)|
|Number of pages||6|
|Publication status||Published - 7 Nov 2017|
|Event||2017 IEEE International Test Conference in Asia, ITC-Asia 2017 - TWTC Nangang Exhibition Hall, Taipei City, Taiwan|
Duration: 13 Sep 2017 → 15 Sep 2017
|Conference||2017 IEEE International Test Conference in Asia, ITC-Asia 2017|
|Period||13/09/17 → 15/09/17|