Negative Testing Revisited – Vehicle Control Systems (Part 2)

If I were to attribute the current issues with automobile systems to a specific cause, I would say that car and truck manufacturers have been affected by a major transition from mechanical engineering to software and systems engineering. However, it does not appear that they have changed their thinking sufficiently to allow for the less tangible aspects of software reliability and resiliency. Somewhat coincidentally, I recently published an article on “Investing in Software Resiliency.” It is in the September/October 2009 issue of CrossTalk magazine and is available at no charge at www.stsc.hill.af.mil/crosstalk/2009/09/0909axelrod.html . While not specifically directed at control systems, the article does raise a number of issues that could be applicable to them. The use of redundant, self-correcting software/firmware is strongly recommended as vehicles become increasingly dependent on electronic control systems. A number of European vehicles (in particular) have systems that cut the power when brake and accelerator are applied simultaneously. For some time now, you have not been able to start a car unless it is in “park” or “neutral.” But, as we find the automobile manufacturers scrambling to incorporate advanced, sophisticated systems that keep the car in lane, slow it down if approaching an obstacle or another car too quickly, or parallel park all by itself, one can only imagine what might happen if such systems fail in a dangerous way. Automakers are trapped into the same tradeoff between features and security/safety as are commercial software vendors.

Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*