Skip to content

Industrial manufacturer · EHS / Manufacturing

Predictive safety AI reduced recordable incidents across 6 plants

38%

fewer recordable incidents in year one

6

plants on one EHS platform

3 wks

from audit request to evidence pack

The challenge

A manufacturer operating six plants tracked incidents in spreadsheets and a legacy tool nobody trusted. Audits took weeks of document hunting, near-misses went unreported, and safety leadership had no way to see risk building before incidents occurred.

Our approach

We deployed the Primis EHS Suite across all sites — incident management, risk assessment, audits, and training in one system — then trained predictive models on three years of migrated incident and observation history to flag elevated-risk sites, shifts, and task types.

What we built

  • Primis EHS Suite rollout across 6 plants with legacy data migration
  • Mobile-first incident and near-miss reporting to lift capture rates
  • Predictive incident-risk models with weekly risk briefings per site
  • Compliance-document agents tracking permits and regulatory updates
  • Live safety dashboards for plant managers and group leadership

Results

  • Recordable incidents fell 38% in the first full year of operation.
  • Near-miss reporting volume tripled — the leading-indicator data the models depend on.
  • Audit evidence packs assembled in under three weeks, down from two months.
  • Predicted high-risk windows now drive targeted toolbox talks and inspections.
"For the first time, we act on where risk is going — not just where it has been."
— Group EHS Director

Related services

Illustrative engagement scenario representative of Primis delivery work, anonymized for client confidentiality.

Get a similar outcome.

It started with one conversation about a bottleneck. Tell us yours.