A.N. Creation Office ยท Project recordPlastic-base snap mousetrap. Example (illustrative figures are marked).

Risk analysis table

Each way it could fail, scored 1 to 10 for how bad it would be (S), how likely (O) and how hard to catch (D); RPN is the three multiplied. Indented lines are the actions taken and the new scores. The most severe are shaded.

idFailure mode / actionseffectSODRPNstatus
F1Base cracks at the spring seatAs proposed, the trap fails. A user can be hurt.845
Add a short rib at the spring seatThe rib makes a crack less likely, and a crack is less severe.735
Use an engineered resinThis is the lowest score available without a new tool concept.63590accepted
F2Nylon swells in the barnThe base, the trigger, and the bail no longer fit. No action was taken. The proposal score stands.538120accepted
F3Base cracks when dropped on concreteAs proposed, the edge cracks. The trap can still be set.456
Thicken the landing edgeFewer drops crack it, and a crack is easier to see.33545accepted
F4Trigger fires with no mouseAs proposed, bait is wasted.447
Increase trip force by 10 percentFalse trips are less common. This is the lowest practical score for this trigger.42648accepted
F5Spring takes a setAs proposed, the trap gets weak after it sits cocked.634
Specify stress-relieved music wireThe spring holds its force.42324accepted
F6Staple pulls out of the nylonAs proposed, the spring comes off the base.533
Barb the staple legThe staple stays in the post.32212below threshold
F7Pedal pin walks outAs proposed, the pedal falls off.533
Head both ends of the pinThe pin cannot walk out.2228below threshold
F8Warning label comes off the nylonAs proposed, the words are gone.333
Use a label adhesive made for nylonThe label stays. The molded mark is still the backup.2228below threshold
F9Blister cracks in the storeAs proposed, the trap is loose in the pack.332
Thicken the blisterThe pack holds together.2214below threshold
F10Bail sticksAs proposed, the trap does not fire.634
Polish the bail contactThe bail releases. Further polish does not help.42324accepted
F11A finger stays in the path of the bail while settingAs proposed, someone can be hurt.935
Mold a set-from-the-side markThe hand is shown where to press.624
Mold a finger guard beside the bailThis is the lowest score for a trap a person still has to set by hand.52330accepted
F12Base warps and the bail bindsAs proposed, the trap will not set.634
Cool the base in the fixture before it is releasedThe base stays flat.42324below threshold
F13Bait will not stay on the pedalAs proposed, the mouse passes the trap.443
Mold a bait spike on the pedalThe bait stays. A pocket was not added.2228below threshold
F14Kill bar wears a groove in the pivotAs proposed, the bar drags after many sets.433
Radius the pivot holesThe bar does not cut a groove in ordinary use.2228below threshold
F15Trigger plate flashes and the pedal sticksAs proposed, the trap does not trip.534
Add the pedal pivot to the trim checkFlash at the pivot is caught. This is as low as the check can take it.32212below threshold
Sum of residual RPNs465
Accepted, the first 80%Ranked by residual score. The running total first passes 372 on F10.381
Below threshold, the last 20%Leftover risk, not a sign-off: F6, F7, F8, F9, F12, F13, F14, F15. The cut is 80% of the risk, not a count of issues. Accepting a row means this is as much as will be spent to lower it. These are not significant. An obvious fix that does not raise the cost is still worth doing.84

Readable copy built by record-view.py from the sheet of the same name. The sheet is the record.