Measure SIF exposure by counting the situations that could cause a serious injury or fatality, not the injuries themselves. For forklifts, the core event is a high-energy interaction: a pedestrian inside a set distance of a moving truck, a truck entering a pedestrian zone at speed, or two trucks closing at a blind corner. Count these events, divide by operating hours, and split the rate by zone, shift, truck and task. A falling rate shows exposure shrinking even in a year with no recordable injuries, which a lagging injury rate cannot show.
The full guide is SIF exposure in forklift operations, which covers precursor events and severity potential. Normalizing matters for the same reason BLS converts injury counts to a rate per 200,000 hours worked: it lets you compare a busy shift with a quiet one.
Data can come from proximity or camera systems that log interactions automatically, or from structured observation where no system is installed. Define severity tiers before you start, for example by truck speed and separation distance, and keep the definitions fixed so trends are real. Review the top zones weekly and record which control changed after each review, so the metric drives fixes rather than reports. See SIF prevention for forklift fleets.