Functional Safety

by Gary Weidner, Editor / Published December 2013

Traditionally, the committees that write product safety standards set about to identify the types of hazards that might be associated with a particular type of product, for example, pressure washers. Then the committees write requirements for protecting against those hazards. Example hazards are electric shock, hot surfaces, fire, sharp edges, moving parts, pressure injection, and so on. But the writers of product safety standards can’t foresee every possible hazard for every version of every product.

Additionally, for some time now, the auto industry has been dealing with a different kind of hazard situation. Suppose a driver jumps on the brakes in a panic, but for some reason the anti-lock braking system kicks in only on one side of the car. That could result in a serious accident. This situation is different from the examples above, because it doesn’t involve “physical” hazards. The hazardous situation arose because of an electrical/electronic control malfunction. The increasing use of electronic controls has spawned a widely spreading safety specialty: “functional safety.”

Functional safety in turn is built in part on what is called “risk assessment.” More and more, it’s required that the manufacturer of a product not just comply with traditional safety requirements, but also perform a risk assessment. The risk assessment, often done by means of a “Failure Modes and Effect Analysis,” has to consider just about every conceivable malfunction of everything from a microchip to a pushbutton to a wire.

Risk assessment and functional safety are gradually spreading their tentacles through one industry after another. And if you want to sell a machine in the European Union, this sort of activity is required by the Machinery Directive.

Perhaps, this is all well and good. What’s interesting and significant is that there seems to be little data to determine how much safer all this work has made products. That’s significant because carrying out functional safety requires a big investment. I’ve seen an estimate that the added development, analysis, design, and validation work can require a 15–30 percent increase in engineering hours. Am I disparaging functional safety? No. But mankind doesn’t have unlimited resources, so some kind of cost-benefit data are sorely needed.

Several standards for functional safety have been published in recent years, with more to come. Manufacturers and component suppliers need to be aware that this won’t go away. It is better to stay alert on the subject than be blindsided.

Gary Weidner
garyw@adpub.com
(800) 525-7038