CANbids

CANbids — CAN-Based Infrastructure for Dependable Systems

Project type: Spanish research project

Period: 2009 - 2011

Focus Area: Design, evaluation, validation, implementation and verification of a fault-tolerant infrastructure based on CAN that could be used to support the execution of highly-dependable distributed control applications.

This work is supported in part by the Spanish Science and Innovation Ministry with grant DPI2008-02195, and in part by FEDER funding.

The Controller Area Network (CAN) protocol is a fieldbus communication protocol that was first devised for in-vehicle control application and that has been widely adopted in many other areas within the distributed embedded control systems field. CAN is nowadays a mature technology whose tremendous success has been mainly caused by its error control features, low latency, network wide bus access priority and real-time response. In addition, CAN’s widespread use has caused the price of its components to drop to some levels where other protocols cannot compete.

Despite these significant advantages, there is an extended belief that CAN is not suitable for critical applications, mainly because of the following dependability limitations: (1) Limited data consistency; (2) Limited error containment; (3) Limited support for fault tolerance and (4) Lack of clock synchronization. Nevertheless, several researchers believe that CAN will be able to support safety-critical applications if these limitations are overcome with the proper enhancements. This possibility is very appealing for many application domains, since CAN components are much cheaper than those of the natural competitors of CAN in highly dependable systems: e.g. FlexRay or TTA. A suitable and specific application for these enhanced CAN-based systems would be critical in-vehicle applications such as X-by-Wire because the use of CAN permits to take advantage of the know-how and expertise that engineering teams of car manufactures have gained in using and programming this technology during the last decades.

Several researchers, including the members of our group, have proposed mechanisms and enhancements intended to overcome the aforementioned CAN dependability limitations. Taking all this previous research as our starting point, the main goal of the present project is to design, implement and validate a CAN-based infrastructure for supporting the execution of highly-dependable distributed control applications.

This infrastructure, which is called CANbids (CAN-Based Infrastructure for Dependable Systems), will use the different mechanisms proposed by our group and by others as building pieces. Given that these mechanisms will be selected among the state-of-the-art in CAN technology, the dependability evaluation of the final product is likely to provide also a clear idea on the maximum dependability that is attainable with the CAN technology. Likewise, since for the highly-dependable infrastructures built on communication technologies different from CAN there are no mathematical dependability assessment results available, this project is also likely to open room for said studies and for their comparison with the results obtained for CANbids.

 

Demostration

Project Leader

Julián Proenza Arenas

Julián Proenza Arenas

Project Collaborators

Related Publications