FMCG — Snacks Manufacturing
Reduction in picker travel distance per shift
Days-of-cover achieved, down from a 35-day target baseline
About
The manufacturer wanted to turn a cramped distribution centre into a model of efficiency. The warehouse sat at 95% cube yet pickers walked 12 km a shift, and static reorder points triggered both stockouts and surpluses. We built a column-generation slotting model and a stochastic EOQ simulator, guiding pickers through handheld scanners in optimal sequence.
Industry
FMCG — Snacks Manufacturing
Company size
1,000 – 5,000 employees
Founded
1991
The Company
A full warehouse that still felt too small
The manufacturer runs a high-volume snacks distribution centre operating at 95% cube utilisation. Management wanted to turn the tight footprint into a genuine efficiency advantage rather than a constant capacity worry.
Despite being nearly full, the warehouse wasn't organised around actual pick frequency — fast-moving SKUs and slow-moving SKUs sat wherever there happened to be space when they arrived.
The challenge
A cramped DC with static reorder points and long walks
Pickers covered roughly 12 km per shift, much of it walking past high-frequency SKUs to reach low-frequency ones stored closer to the pick face simply because of when they'd been slotted in. Static reorder points, set once and rarely revisited, meant the warehouse was simultaneously stocking out on fast movers and sitting on surplus of slow ones.
The combination — inefficient layout plus stale reorder logic — meant the DC was working harder than it needed to for the throughput it was actually achieving.
The Solution
Column-generation slotting and a stochastic reorder engine
We built a column-generation slotting model that re-laid the pick-face around actual pick frequency rather than arrival order, paired with a stochastic EOQ simulator that recalculated reorder points against real demand variability instead of a static rule of thumb. Handheld scanners now guide pickers through the warehouse in optimised sequence.
The two systems work together: the slotting model determines where a SKU should sit, and the reorder engine determines how much of it should be there at any given time — solving the layout and the inventory-level problem as one connected decision rather than two separate ones.
The Results
28% less walking, 34 days of cover, Rs. 3.4 Cr freed
Picker travel distance fell 28%, well past the original 20% target, as fast-moving SKUs moved to the most accessible pick locations. Days-of-cover settled at 34 days, inside the 35-day target, as reorder points were recalibrated to actual demand patterns rather than historical guesswork.
Rs. 3.4 crore in working capital was released as excess safety stock built against stale reorder points came down to levels the actual demand variability justified.
KEEP READING
Explore more studies on AI, analytics, and enterprise intelligence.
