Warehouse Management

Inventory management says how much you have. Warehouse management says exactly where it is and how it moves.

When you need a WMS

Signal Implication
Stock tracked at storage-location level only Basic inventory management is enough
You need bin-level location WMS
Directed put-away and picking WMS
Multiple picks per trip, wave/batch picking WMS
Handling units / pallets / licence plates WMS
RF scanners or voice picking WMS
3PL operations, billing for storage/handling WMS
Automated equipment (AS/RS, conveyors, AMRs) WMS + WCS

Small operations run ERP inventory only. Beyond a few thousand SKUs and a few pickers, a WMS pays for itself in labour alone.

The warehouse structure 

Warehouse
 └── Storage type       (racking, bulk, cold store, hazmat, staging, returns)
      └── Storage section  (fast-moving zone, slow zone)
           └── Storage bin   A-01-02-3  (aisle-rack-level-position)
                └── Quant     the actual stock in that bin

A quant is the atomic unit: material + batch + stock category in one specific bin, with a quantity.

Handling units and licence plates

A handling unit (HU) / licence plate (LPN) is a physical package with its own barcode:

HU 00340123456789012345 (SSCC)
 ├── 24 × Product A, batch B2601
 ├── 12 × Product B, batch B2599
 └── packed on pallet EUR1, gross weight 340 kg

Scanning one label moves everything inside. Nested HUs — cartons on a pallet, pallets in a container — are standard. The SSCC (Serial Shipping Container Code, GS1) is the global identifier and is what appears on the GS1-128 shipping label.

Inbound process 

ASN (Advance Shipping Notice) received  ──▶ expected receipt created
              │
        Truck arrives → dock appointment
              │
        Unload → count → scan
              │
        Goods receipt posted (ERP: stock + accounting)
              │
        Quality inspection (if required) → hold or release
              │
        PUT-AWAY: system determines destination bin
              │
        Confirm put-away → stock now in a specific bin

Put-away strategies

Strategy Rule
Fixed bin Each material has a home bin
Next empty bin First available
Near picking bin Minimise travel to the pick face
Addition to existing stock Consolidate with the same material/batch
Bulk storage Full pallets to bulk area
Pallet type / capacity check Only where it physically fits
Zone by velocity Fast movers near the dispatch dock
Chaotic / random System remembers; humans don’t need to. Maximises density

Constraints the strategy must respect: weight limits, height, hazmat separation rules, temperature zones, food/non-food separation, and mixing rules (can two batches share a bin?).

Outbound process 

Sales order ──▶ Delivery created ──▶ Wave/release
                                        │
                          Pick task generation (bin, qty, route)
                                        │
                          PICKING (RF, voice, pick-to-light, paper)
                                        │
                          Staging → PACKING → HU creation → label
                                        │
                          Loading → Goods issue posted (stock ↓, COGS)
                                        │
                          Shipment → ASN sent to customer

Picking strategies

Strategy How Best for
Discrete / order picking One picker, one order, start to finish Small operations, large orders
Batch picking One picker, many orders, one trip Many small orders
Zone picking Picker owns a zone; order passes between zones Large warehouses
Wave picking Release a batch of orders together by criteria (carrier cut-off, route) Scheduled dispatch
Cluster picking Multiple order totes on one cart E-commerce
Goods-to-person Robots bring shelves to a stationary picker High-volume automation

Stock removal strategies

Strategy Rule
FIFO Oldest goods receipt first
FEFO First Expired, First Out — mandatory for pharma and food
LIFO Newest first — rare, sometimes for bulk piles
Partial quantities first Empty part-pallets before breaking full ones
Fixed bin Always from the pick face
Large/small quantity split Full pallets from bulk, cases from the pick face

FEFO is a compliance requirement, not an optimisation. If you build for food, pharma, or chemicals, FEFO with expiry-date checking is table stakes, along with a minimum-remaining-shelf-life rule per customer (some retailers reject stock with less than 2/3 of its life left).

Replenishment

Moving stock from bulk storage to the pick face before it runs out:

  • Min/max — when pick face drops below min, replenish to max
  • Demand-driven — replenish for the next wave’s requirements
  • Order-driven — triggered by a specific pick that will empty the bin

If replenishment fails, picking stops. It’s a classic hidden bottleneck.

Cross-docking

Goods go from inbound dock straight to outbound without being put away:

  • Opportunistic — the system notices an inbound receipt matches an open order
  • Planned — inbound is pre-allocated to outbound before it arrives

Saves put-away and picking labour entirely. Requires tight timing and good ASN data.

Task and labour management 

warehouse_task(id, type,        -- PUTAWAY | PICK | REPLENISH | COUNT | MOVE
               source_bin, dest_bin, material, batch, quantity, handling_unit,
               priority, assigned_to, status, queue,
               created_at, started_at, confirmed_at,
               planned_duration, actual_duration)

Tasks are dispatched to operators by queue, zone, and priority. Interleaving (combining a put-away with a nearby pick on the same trip) can cut travel time by 20–30%.

Labour management adds engineered standards: expected time per task type and distance, actual vs standard performance per operator, and productivity reporting. Sensitive to implement — it’s performance monitoring of individual workers, with the employee-relations and (in some jurisdictions) legal implications that carries.

Counting

  • Cycle counting — continuous rolling counts by ABC frequency
  • Zero-count — count when a bin is emptied by a pick. Nearly free
  • Blind counting — the counter doesn’t see the expected quantity. Better data, slower
  • Bin-level accuracy is the WMS metric; item-level accuracy is the ERP metric. They’re different and both matter

Integration with ERP

The boundary is the single most important design decision:

ERP                                  WMS
───                                  ───
Purchase order         ──────────▶   Expected receipt
                       ◀──────────   Receipt confirmation (qty, batch, HU)
Goods receipt posting

Delivery/shipment      ──────────▶   Outbound order
                       ◀──────────   Pick confirmation, packed HUs, weights
Goods issue posting

Stock transfer         ──────────▶   Move task
                       ◀──────────   Confirmation

                       ◀──────────   Adjustment (count difference)
Adjustment posting

Rules for this integration:

  1. The ERP owns the financial posting. The WMS confirms physical reality; the ERP books it.
  2. Idempotency on every message. Warehouse networks drop connections; messages get retried.
  3. The WMS must work offline for a while. Warehouses cannot stop because the ERP is down for patching. Queue and replay.
  4. Reconcile daily. WMS quantity per material/plant must equal ERP quantity. Publish the differences every morning.
  5. Decide who owns batch determination — usually the WMS, since it knows what’s physically in which bin.
  6. Handle partial confirmations. A pick of 100 that yields 97 must flow back correctly and adjust the delivery.

Automation

Technology Role
Barcode / GS1-128 / QR The baseline
RFID Bulk reading, apparel retail, high-value tracking
Voice picking Hands-free, good accuracy
Pick-to-light / put-to-light High-speed zones
AS/RS Automated storage and retrieval cranes
AMR / AGV Mobile robots
Conveyor + sorter High-volume parcel
WCS / WES Warehouse Control/Execution System — the real-time layer between WMS and the machines

The WCS boundary matters: WMS decides what to do, WCS decides how the machine does it. Never put millisecond-level equipment control in the WMS.

Metrics

Metric Meaning
Lines picked per hour Productivity
Pick accuracy Errors per thousand picks
Dock-to-stock time Receipt to available
Order cycle time Release to shipped
OTIF On time in full
Inventory accuracy (bin level) Data quality
Space utilisation Density
Cost per order/line Economics

Next