ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
ISSUE 01 — MAY 2026
01 / 15
ORACLE WMS CLOUD · RELEASE 26B · OUTBOUND SERIES

THE WAVE
MACHINE

Wave planning is the heartbeat of outbound fulfilment. This issue dissects every field, algorithm, and integration that makes Oracle WMS Cloud 26B tick.

WAVE PLANNING & PICKING
OUTBOUND DOCK Wave path Cluster path
ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
ISSUE 01 — MAY 2026
02 / 15
PRIMER · ORACLE FUSION WMS CLOUD

What
is a
Wave?

In Oracle Fusion WMS Cloud, a wave is a batch of outbound orders grouped and released to the warehouse floor as one unit of work. The system bundles dozens — or hundreds — of orders, then simultaneously allocates inventory, sequences pick paths, and generates tasks for every operator.

Think of it as a conductor's downbeat: the moment a wave fires, the WMS engine runs allocation, cubing, task creation, and pick-slip grouping — all in one pass.

Batch
Many orders processed simultaneously — not one-by-one
Allocate
Inventory reserved and pick tasks generated in one engine pass
Release
Workers receive RF-directed tasks the instant the wave fires
WAVE LIFECYCLE IN FUSION WMS CLOUD 1 Orders Queued Outbound orders released from Oracle Fusion OMS / SCM into WMS 2 Wave Template Fires Allocation method, cubing mode & pick-slip grouping rule applied 3 Inventory Allocated Stock reserved; OBLPNs (outbound containers) created and cubed 4 Pick Tasks Generated RF-directed tasks sent to pickers; replenishment tasks auto-added 5 Pick, Pack & Manifest Workers execute tasks; OBLPNs packed, verified & staged for ship 6 Ship Confirmed → Fusion ASN sent to Fusion Inventory via OIC pre-built integration flow
ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
THE PROBLEM
03 / 15
THE PROBLEM WITH NAIVE FULFILMENT

Without a wave, the warehouse is a pinball machine.

Order-by-order release sends workers on overlapping paths, creates congestion at hot locations, and leaves pack stations idle while the next order is still being picked.

Wave planning batches demand into coordinated releases — picks, replenishments, packing, and manifesting finish together. Oracle WMS Cloud 26B supports multiple picking task modes — Order, Wave, Cluster, and Zone — configured via Task Creation Templates within the wave setup.

The scheduler runs on-demand or on a timed interval — no middleware, no batch job, no external trigger required. Source: Oracle WMS Cloud 26B Impl & Config Guide G52815-01

WHAT GOES WRONG WITHOUT WAVES
40%
travel time wasted on overlapping pick paths when orders release individually
industry estimate
3×
more OBLPNs created when cubing runs per-order instead of across a wave
industry estimate
24D+
Oracle WMS release that introduced Constraint Model Cubing — available in 26B and all subsequent releases
10
pre-built OIC flows to Fusion Inventory — zero custom integration code
ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
THE ARCHITECTURE
04 / 15
THE WAVE ENGINE

The wave template is the DNA of every fulfilment run.

Six fields in one configuration object control how Oracle WMS Cloud selects inventory, groups picks, dimensions cartons, and routes operators. Every future wave inherits them exactly.

Allocation Method which inventory lines to pull
Allocation Mode reserve first, then active
Cubing Mode how many cartons to build
Task Creation Template what type of RF task to create
Pick Slip Grouping Rule how many lines per operator
Manifest OBLPN by Order merge or split outbound cartons
SHIPMENT REQUESTS from Oracle Fusion SCM WAVE PLANNER Template · Order Selector · Scheduler ALLOCATION ENGINE Method · Mode · Reserve → Active CUBING ENGINE Constraint Model (24D+) · OBLPN build TASK + PICK SLIP Grouping Rule · RF dispatch PACK + MANIFEST OBLPN by Order · carrier label SHIPMENT CONFIRM OIC → Fusion Inventory update
ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
THE ARCHITECTURE
05 / 15
PICKING TASK PATTERNS — ORACLE WMS CLOUD 26B

One wave engine. Four ways to move product to the dock.

Order Picking

One picker completes all lines for a single order before moving on. High accuracy, ideal for low-volume or B2B.

Mode: Direct
Note: Fusion Inventory also has a native Pick Wave for managed orgs — distinct from WMS Cloud wave engine.
Wave Picking

Lines across many orders release simultaneously, line by line. High throughput for uniform SKU profiles — the default mode.

Mode: Reserve→Active
O1O2O3
Cluster Picking

One picker carries a multi-slot tote and sorts picks into order compartments in real time. Eliminates a sort pass for B2C.

Task Type: Cluster Pick
AB
Zone Picking

Workers fenced to a zone see only their area. Release 25D adds Zone Picking without Drop Location via RF Pack NC Active.

Param: Zone w/o Drop

Source: Oracle WMS Cloud 26B Impl & Config Guide G52815-01 Wave Planning

ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
THE ARCHITECTURE
06 / 15
WAVE TEMPLATE — 6 DECISION FIELDS

Six fields set at rest. A thousand outcomes in motion.

The Wave Template is a ruleset, not a schedule. Once saved, every wave referencing it inherits identical allocation logic, cubing behaviour, and task dispatch rules.

A facility can maintain multiple templates — small-parcel B2C with Constraint Model cubing, bulk pallet B2B without cubing, B2B cluster picking — all coexisting, selected by order type at wave release time.

// navigation path in WMS Cloud UI
Configuration → Outbound → Wave Templates
Wave Template · SMALL-PARCEL-B2C ALLOCATION METHOD Outbound Order Lines standard ALLOCATION MODE Reserve First, then Active reserve→active CUBING MODE Constraint Model ★ New in 24D constraint_model TASK CREATION TEMPLATE PICK-ACTIVE-PARCEL pick_active PICK SLIP GROUPING RULE BY-ZONE-20-LINES zone · max 20 MANIFEST OBLPN BY ORDER Yes — group OBLPNs by order true Save Cancel
ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
THE ARCHITECTURE
07 / 15
CONSTRAINT MODEL CUBING — ORACLE WMS 24D (F35043)

Fewer cartons. Same product. The algorithm does the geometry.

Classical cubing assigns lines sequentially and overflows to a new box. Oracle's Constraint Model re-solves the packing problem across all lines simultaneously. The example below is from Oracle 24D What's New doc F35043.

BEFORE — STANDARD CUBING
OBLPN 1 80% OBLPN 2 90% OBLPN 3 20% wasted
3 OBLPNs
~30% wasted capacity · 3 carrier labels
24D
algo
AFTER — CONSTRAINT MODEL CUBING
OBLPN 1 100% OBLPN 2 90%
2 OBLPNs
Overflow redistributed · 1 fewer label

Source: Oracle WMS Cloud 24D What's New F35043 — Constraint Model Cubing section

ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
THE ARCHITECTURE
08 / 15
THE PICK SLIP GROUPING RULE

The pick slip grouping rule is the unit of work handed to one operator.

After allocation and cubing, Oracle WMS Cloud groups picks into pick slips. The Pick Slip Grouping Rule — confirmed in both Oracle WMS Cloud and Fusion Inventory Management documentation — determines which allocations share a slip and how far one RF operator travels before returning to a consolidation point.

Too many lines per slip: operators block aisles. Too few: system overhead dominates cycle time.

Group by Zonezone_id
Max Lines per Slip1 – 999
Group by OBLPNoblpn_id
Group by Order + Zonerecommended
Group by Carrier / Routecarrier_route
Runtime: wave → pick slips → operators
WAVE RELEASE Allocation Pool (all wave lines) Pick Slip Grouping Rule Slip #001 Zone A · 18 lines Slip #002 Zone B · 20 lines Slip #003 Zone C · 15 lines OP1 OP2 OP3 RF tasks dispatched simultaneously
ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
INTELLIGENCE LAYER
09 / 15
ORACLE WMS 26B — AI AGENTIC COMMAND CENTER

The command center that surfaces today's at-risk orders before they miss the dock.

Oracle WMS Cloud 26B introduces the Logistics Execution Command Center — an AI Agentic App consolidating WMS and Oracle Transportation Management (OTM) data into a single view. Planners use natural language ("Ask Oracle") to surface at-risk orders and run waves directly from the dashboard. The predictive module adds order cycle time, processing time, and waiting time forecasts.

Unlike traditional dashboards needing custom queries, the Command Center ingests WMS and OTM data directly — auto-detecting at-risk orders, surfacing priority actions, and enabling one-click wave execution for the orders most at risk of missing carrier cut-off. Source: Oracle WMS Cloud 26B What's New.

THREE CONFIRMED 26B CAPABILITIES
① At-risk orders surfaced — orders within 3 days ranked by fulfilment risk
② Run wave for at-risk orders directly from the dashboard
③ Predicted cycle time, processing time & waiting time per wave
Logistics Execution Command Center 26B What's New · Confirmed Cycle Time (est.) 2.4h Volume Forecast 1,840 Shortfall Risk HIGH PREDICTED WAVE THROUGHPUT · NEXT 8H 08:00 09:00 10:00 11:00 12:00 13:00 14:00 15:00 forecast ▶ ⚠ SHORTFALL ALERT SKU A2847-BL insufficient for Wave #14 at 14:00. Suggested: advance replenishment task now. PICK FACE HEALTH Zones A–C: OK Zone D: Low Zone E: Critical Source: Oracle WMS Cloud 26B AI/ML Guide · Predictive Fulfilment Dashboard
ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
INTELLIGENCE LAYER
10 / 15
REPLENISHMENT WITHIN THE WAVE CYCLE

A wave that runs out of pick face stock is a wave that stalls.

Oracle WMS Cloud generates replenishment tasks as part of wave processing — not a separate batch job. When the allocation engine finds a pick face below its minimum, it auto-creates a replenishment work order queued ahead of the picks that need it.

In a correctly tuned warehouse, forklift operators receive replenishment RF tasks at the same time pick operators receive pick slips — parallel, not series.

1
Min/Max on Pick Face
Each active pick location has a minimum quantity. Allocation engine checks it before assigning the line.
2
Auto Replenishment Task
Wave processing creates a Replenishment task dispatched to forklift RF queues automatically.
3
Parallel Execution
Replenishment and pick tasks run concurrently. Picks for depleted locations are held — not cancelled — until replenishment confirms.
Wave cycle — replenishment runs in parallel
Wave Processing Replenishment Pick Tasks A–C Pick Tasks D (held) Pack + Manifest held T+0 T+30m T+60m replen done
ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
FUSION INTEGRATION
11 / 15
ORACLE INTEGRATION (OIC) — WMS TO FUSION INVENTORY

Ten pre-built flows. Minimal custom integration code.

For teams running Oracle Fusion SCM end-to-end, Oracle ships 10 prebuilt Oracle Integration (OIC) accelerator flows between WMS Cloud and Fusion Inventory Management. Each is a prebuilt OIC accelerator recipe — configuration and setup are still required, but the integration logic is Oracle-provided.

This is not a bridge between foreign systems. It is the native plumbing Oracle built for its own stack — when a wave confirms a shipment, the flow fires and Fusion Inventory updates in the same run.

Source document
Oracle SCM Cloud Integration Playbooks 26B
Section: WMS Cloud Integrations
10 OIC PREBUILT FLOWS
01
Inventory Transactions
02
Receipt Advice PO / RMA
03
Supplier ASN
04
Receipt Confirmation
05
Shipment Request
06
Update Shipment Request
07
Order Lock / Unlock
08
Shipment Confirmation
09
Backorder for Shipment
10
Receipt Advice for RMA

Source: Oracle SCM Cloud Integration Playbooks 26B

ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
FUSION INTEGRATION
12 / 15
WAVE-TO-SHIPMENT — OIC FLOW SEQUENCE

The integration is the fulfilment handshake. Every wave confirms twice.

Oracle Fusion Inventory Management sends a Shipment Request to WMS Cloud via OIC. After picks complete, WMS fires a Shipment Confirmation back — Fusion Inventory updates on-hand quantities, lots, serials, and source document status. No manual step, no file drop. Source: Oracle Fusion Cloud SCM: Using Inventory Management, Ch. 13.

FUSION SCM OIC FLOWS WMS CLOUD Sales Order Fusion SCM Shipment Req Flow #05 Outbound Order WMS Cloud Wave Release Allocate + cube Pick + Pack RF tasks complete Ship Confirm WMS manifest Shipment Confirm Flow #08 Inventory Update Fusion Inventory Sales Order Close Fusion SCM
ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
FIELD NOTES
13 / 15
FROM 25C ZONE PICKING WITHOUT DROP LOCATION

The drop location disappears. The pick does not.

Classic zone picking required operators to carry picked goods to a fixed drop location before the next zone's worker could continue — creating systematic congestion at the drop point.

Oracle WMS Cloud documents Zone Picking without Drop Location (from Release 25C) via the RF Pack NC Active module. Workers scan the OBLPN at the pick face and proceed to the next line directly.

CONFIGURATION PARAMETER
Zone Picking without Drop Location
Module: RF Pack NC Active · Release: 25D+
BEFORE 25C — WITH DROP LOCATION
ZONE A DROP congestion ZONE B waiting
AFTER 25D — ZONE PICKING WITHOUT DROP
ZONE A scan OBLPN in-aisle direct ZONE B already picking
ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
FIELD NOTES
14 / 15
FROM THE IMPLEMENTATION FLOOR — WHAT GOES WRONG

The wave is sound. The configuration is where implementations fail.

MISTAKE 01
One wave template for every order type.

A single template tries to satisfy small-parcel B2C, bulk pallet B2B, and cold-chain simultaneously. Cubing creates oversized cartons for B2B while B2C cluster picks time out.

THE FIX

One template per order-type segment. Use the Order Selector to route each type to its template at wave release.

MISTAKE 02
Enabling Constraint Model Cubing before container types are defined.

The algorithm needs a container type hierarchy with dimensions and weight limits. Without it, cubing optimisation cannot run effectively. Set up Container Types first and test in a sandbox — this is the most common cause of cubing producing no improvement.

THE FIX

Complete Container Types setup first. Test in sandbox before enabling in production.

WATCH OUT
Leaving replenishment minimums at zero before go-live.

If pick face minimums are zero, the wave allocates directly from reserve storage — skipping active pick faces. Picks travel to reserve locations, destroying cycle time during the first live waves.

THE FIX

Set min/max on all active pick locations as part of go-live data prep — not as an afterthought.

Can I mix cluster picking and zone picking in the same wave? Typically, no — each template is built for one task pattern. Create separate templates and release concurrent waves against different order selectors. Oracle does not document a hard restriction, but mixing task types in one template creates conflicting allocation logic in practice.

Pick Slip sizing rule of thumb: Start at 15–20 lines per task for mixed-SKU warehouses (implementation rule of thumb). Increase only after measuring operator task completion time per slip — not before.

ORACLE WMS WAVE NOTES · A CREATIVE FLAKES PUBLICATION
VERDICT
15 / 15
THE VERDICT — ISSUE 01

The wave is not a scheduling convenience — it is the heartbeat of the warehouse, and Oracle 26B finally gives you the instruments to read it.

Six fields. Ten integration flows. One AI dashboard. Three of the biggest constraints in outbound fulfilment — carton waste, zone congestion, and integration latency — have documented solutions in the current release. The question is not whether the platform supports it. The question is whether the implementation plan does.

26B
current release
6
template fields
10
OIC prebuilt flows
0
custom code needed
ORACLE WMS WAVE NOTES — ISSUE 01 · MAY 2026
A Creative Flakes Publication · creativeflakes.com
Sources: Oracle WMS Cloud 26B G52815-01 · 24D What's New F35043 · SCM Integration Playbooks 26B · 26B AIML Guide