Home > Research > Publications & Outputs > Reconfigurable circuits for fault tolerant systems

Associated organisational unit

View graph of relations

Reconfigurable circuits for fault tolerant systems: factors to consider

Research output: Contribution to conferenceConference paper


Publication date21/06/2002
Number of pages4
<mark>Original language</mark>English


Workshop8th IEEE Internationa Mixed Signal Testing Workshop


In safety critical applications of advanced mixed signal technology, such as in automotive systems, the concept of zero-failure may be implemented by integrating failure management functions based on self-test, online monitoring, failure diagnostic and failure mode compensation in parallel with highly robust design style and packaging technology. Here the primary objectives of
advanced test techniques are the detection of conditions which may lead to systems failure, and the initiation of appropriate action to ensure safety while maintaining system performance.
Traditionally, redundant systems for fault tolerance are used in safety critical applications, such as nuclear power stations and aerospace. These systems are not economical for the majority of next generation safety critical and high
reliability applications as it is impractical to implement entire redundant subsystems. Automotive systems need to be produced cheaply and achieve high reliability. In failsafe systems additional circuitry will be required, the extent of which will depend on the architecture of the system and knowledge of the failure effects.
In the event of a potential failure, some kind of backup method has to be provided to ensure a certain reduced system functionality without complete system breakdown.
In the majority of cases this may be achieved through circuit reconfiguration.
This paper identifies the factors involved in designing reconfigurable circuits and systems. Some initial ideas are presented for a typical sensor interface circuit including potential reconfiguration and failure mode compensation techniques which may be capable of achieving improved system fault tolerance.