Work

Case study deck · 9 slides

Paints & coatings

Manpower norms built from the factory floor, not the org chart

A unit of a leading global paints & coatings player

Audit & TeardownProcess Redesign & BuildAdoption & Enablement
01 / 09

The challenge

What was broken

There was no benchmark for how long a changeover should take or how many hands a batch needed, so every headcount conversation was an argument between two guesses. The harder problem was honesty: a floor that knows it's being watched runs by the book, and existing norms had only ever been checked against a calm Tuesday day shift, never a night handover or a short-crew week.

Man-hour cost per batch (indexed), eight months pre-engagement, no activity benchmark, no visible trend to act on
02 / 09

The approach

Scored on the Durable AI Index

We spent time on trust before tracking a single task, looping supervisors in early so they had a stake in the numbers being accurate. Once that held, we logged task-level activity across roughly 80% of running batches to derive draft norms per activity. Because a norm built on average conditions breaks on a bad day, we simulated every draft against night handovers, batch-size swings and short-crew shifts before anything went live.

DURABLE AI INDEXImpact84Feasibility77Stickiness8341at startafter redesign
High impact, low stickiness. The gap was the workflow, not the model.
03 / 09

What we built

The system, not just the model

We redesigned the workflow around the model, then built the pieces that make it run every day.

  1. 1

    Activity-wise task log covering ~80% of batches, by role and station

  2. 2

    Manpower norms per activity, derived from observed task time

  3. 3

    Shift/edge-case simulation against handovers and short-crew scenarios

  4. 4

    SOP and training rollout translating norms into a floor staffing guide

04 / 09

Inside the build

Manpower norm build & validation timeline

How activity-level floor observation turned into shift-tested manpower norms, from first walk-through to plant recognition.

Batches tracked: ~80%Stations: 5 coreSimulated: Night/rush/short-crew

Tools & systems

manual task-time capture (unstaged)activity-benchmark modelshift/edge-case simulationSOP + training rolloutsupervisor co-validation loop
  1. Floor trust-building

    Consultants walked the line with supervisors, explaining the purpose of observation before recording a single task.

  2. Unstaged task tracking

    Logged task time and role coverage across roughly 80% of running batches without altering shift routine.

  3. Norm derivation

    Built per-activity manpower norms from observed task time across mixing, filling, packing, and QC.

  4. Shift & edge-case simulation

    Ran draft norms against night handovers, mid-run batch-size changes, and short-crew scenarios before rollout.

  5. Training & SOP rollout

    Converted validated norms into shift-level SOPs and ran supervisor-led training across all shifts.

  6. Internal recognition

    The unit received an internal award from Asian Paints for cost optimization following the sustained man-hour reduction.

05 / 09

Results

What good looks like

Man-hour costs down

Significant reduction through data-backed manpower reform

80%

Of batches tracked at task level to build the norms

Validated

Norms simulated across shifts and exception scenarios before rollout

Internal award

Recognized by the parent company for excellence in cost optimization

BATCHES WITH TASK-LEVEL TRACKING0%Before80%During the study
06 / 09

How it stuck

Adoption is the deliverable

The norms held because supervisors had seen the raw data and argued the edge cases before rollout, so there was no credibility gap to close. Six months in, the plant used the same benchmarks to catch a new packing-line bottleneck on its own, without calling us back.

050100month 6This engagementTypical pilot
Norm adherence on the floor, first 12 months: typical staffing-reform rollout vs. this engagement
07 / 09
We stopped guessing at headcount and started staffing to what the batch actually needed. That's the part that stuck.

Plant Manager, Production Unit

08 / 09

Want this for your business?

Book a free 30-minute AI opportunity assessment. You will leave with at least one concrete idea.

09 / 09