1999 - 2000 - 2001 - 2002 - 2003 - 2004 - 2005 - 2006 - 2007 - 2008 - 2009 - 2010

Announcement - Agenda - Paper Presentations

Traditionally, safety critical systems have been constructed from low volume hardware and software components specifically designed for safety critical systems. These systems are also typically closed systems, and the exact designs and source code are not available for analysis or comparison; we can only evaluate the reliability based on published reports of accidents attributed to the control systems. However, many of the failures that could have caused catastrophic accidents, but, by luck, only resulted in delays or interruptions in service, can be quietly fixed and not reported. It is the belief of the authors, that safety critical systems can be created from high volume general purpose COTS hardware and open source components and be just as reliable as traditional low volume hardware and closed source software components. We are in fact working with the light rail system in Guadalajara Mexico, and have an agreement to develop various projects, including a monitoring/signaling system using open source software licensed under the GNU General Public License (GPL). So far, we have implemented two GPL licensed systems, one to monitor the a pumping station and generate alarms, and another to monitor the departures at the four extremes of the current light rail system. The second should be classified as a safety critical system, since, operators will become dependent on it to maintain the spacing of trains and thus the safety of the system.

Techniques such as n-version programming, internal consistency checks, design patterns, peer review, etc, can be used in conjunction with COTS and free software components. We can also develop completely different systems that work in parallel to guarantee safety, with hardware voting.

In developing countries that do not have the resources to pay for traditional proprietary safety critical systems, the choice is often between free software / COTS system, or no system upgrades at all. In the case of the light rail system in Guadalajara Mexico, the safety can be greatly improved over the existing manual system using affordable free software and COTS components. In countries like Mexico that need so many things such as improved public health, improved traffic systems and roads, better education systems, it is not reasonable to pay in one area for extremely expensive commercial systems based on standards from rich industrialized countries, when other projects that are possibly more important are left completely without funding. In fact, overall, more lives can be saved by more appropriate allocation of resources.