Machine Design

Switch faults, worker crushed

I'd be interested in your thoughts on the design of controls that fail in such a way as to not cause or increase a hazard.

Edited by Lawrence Kren

I am currently working with lawyers on a productliability case. The piece of equipment in question has an inductively activated SPST sensor switch. The presence of a signal indicates one condition, and the absence of the signal, the other.

When activated, the equipment moves, creating a hazard for the operator. In this particular accident, the wiring harness to the sensor was damaged, so no signal reached the control system. The operator was crushed to death.

The question is, should there not have been two signals, one for each condition? Then, if both signals were absent, the control system could detect a malfunction and not automatically reposition the equipment. In some instances, this could be accomplished with a DPDT switch or two separate switches. This is a small expense to save a man's life.

What do you think?

- Jim Thompson
The Woodlands, Tex.

You really don't give me much to work with, but in general:

1. The control system should be designed such that any inadvertent operation of the machine could not cause injury. If that is not possible, then the operator should be protected by guarding.

2. Any time an operator is placed in harm's way doing service and maintenance, the machine must be locked or tagged out.

It seems to me that the use of fail-safe controls to protect a worker addresses the wrong problem. Any control system eventually can fail. The thing to focus on is, when the control system fails in some manner, the operator cannot be injured.

The equipment manufacturer should have performed a hazard analysis. Had the hazard analysis identified a switch malfunction as creating a hazard, the problem could have been addressed in design.

If I were acting as an expert witness for the plaintiff, I would focus on the deficiencies with the design that allowed the operator to be in harm's way during the normal and expected motion of the machine, not on the operating controls. I hope that this helps.

Lanny Berke is a registered professional engineer and Certified Safety Professional involved in forensic engineering since 1972. Got a question about safety? You can reach Lanny at [email protected].

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.