Warehouse and Intralogistics Optimization With Real-Time Data
DBR77 Team7 min read
On this page
Lines wait for material when the warehouse works from yesterday's plan and cannot see what the line uses today. Optimizing intralogistics with real-time data means connecting the WMS to production: each order release creates a pick, consumption is booked as it happens, and line-side stock below a set minimum creates a replenishment task with a deadline. The goal is fewer surprises at the line.
Warehouse teams are usually measured on picks per hour and stock accuracy. The line judges them on one thing: was the material there when the order started?
Why the line waits when the stock is in the system
The same reasons come up again and again:
- Wrong location. The system shows 12 pallets. Two of them stand in a staging lane nobody scanned.
- Booked too late. Consumption is backflushed when the order closes, so the stock figure runs hours behind the line.
- On hold. The lot is waiting for receiving inspection, but the planner sees it as available.
- Wrong timing. The order sequence changed after a breakdown, and the warehouse is still picking for the old plan.
Each one is a gap between the warehouse record and what happens on the line and in quality, so the warehouse cannot fix it alone. Toyota describes just-in-time as making only what is needed, when it is needed, and in the amount needed. The same rule applies to material delivery, and it only works when the warehouse knows "when" and "how much" as the shift unfolds, not the next morning.
- Release or resequenceCreate or reorder picks
- Material consumedUpdate line-side and warehouse stock
- Line-side stock below minReplenishment task for the route driver
- Changeover plannedBring next material, take back leftovers
- Lot on hold or rejectedBlock the lot, pick an approved one
- Long stop on a linePause deliveries to that line
A released or resequenced order creates or reorders picks, ready 60 minutes before start.
What the warehouse needs to know from production
Six signals cover most needs. Each should trigger a defined action with a time limit.
| Signal | Source | Warehouse action | Typical time limit |
|---|---|---|---|
| Order released or resequenced | Plan, MES | Create or reorder picks for the next orders | Picks ready 60 minutes before order start |
| Material consumed, including scrap | MES, machine counts | Update line-side and warehouse stock | At each pallet or bin change, not at order close |
| Line-side stock below minimum | Bin scan or count | Replenishment task for the route driver | Delivered within 30 minutes |
| Changeover planned | Plan, MES | Bring the next material, take back leftovers | At the line 15 minutes before changeover |
| Lot on hold or rejected | Quality | Block the lot, pick an approved one | Immediately |
| Long stop on a line | MES, machine data | Pause deliveries to that line | When the stop passes 30 minutes |
Most plants already have the data for the first two rows, but it reaches the warehouse too late.
Pull instead of push: supermarkets, kanban and routes
Lean plants solved line supply long before software. The Lean Enterprise Institute defines a supermarket as the location where a predetermined standard inventory is kept to supply downstream processes. When an item is taken, a signal goes back to bring or make more. That signal is a kanban, which the same lexicon describes as a device that gives authorization and instructions for the production or withdrawal of items in a pull system.
Real-time data makes the signal faster and harder to lose. An empty bin scan, or a consumption count from the machine, can be the kanban: it creates a task for the next route without anyone picking up the phone.
Routes matter as much as signals. The Lean Enterprise Institute describes milk runs as a way to move material between facilities and notes that they are similar in concept to material handling routes within facilities. A driver who runs a fixed loop every 30 minutes, with a task list that updates from consumption, usually beats a forklift that answers calls.
Before you move racks, change routes or add an AMR, test the change first. Our DT article on how to simulate intralogistics and material flow shows how.
An illustrative example from an assembly plant
Illustrative example: a plant builds control cabinets on three assembly lines, supplied from one central warehouse by two route drivers. At 09:20, a test bench fails on line 3, and the planner moves a rush order on line 2 forward to 10:00.
Without connected data. The warehouse is still picking for the old sequence. The line 2 lead notices the missing parts at 09:50 and calls the warehouse. One component lot turns out to be on hold for receiving inspection. The order starts at 10:40.
With connected data. The resequence creates new picks at 09:21. The warehouse sees at once that one lot is on hold and picks an approved lot instead. The route driver delivers at 09:52, and leftover material from the paused order goes back on the return trip. The line starts the order at 10:00.
The forklift and the drivers are the same in both versions.
What to measure
Track a few measures that the line will recognize:
- Minutes lines waited for material, per line and per cause: wrong location, late pick, hold, plan change.
- Picks ready before order start, as a share of all orders.
- Replenishment tasks closed within their time limit.
- Location accuracy from cycle counts, by zone.
- Emergency picks per shift: requests that came by phone instead of as a task.
Review them weekly with planning and the line leads. Many material waits start in the plan itself, which is the topic of Production Planning vs Reality: Why APS Plans Fail.
How this works in DBR77 IRIS
DBR77 IRIS runs the warehouse in IRIS WMS, on one data model with production, quality and planning. Locations are modeled from warehouse and zone down to aisle, rack and shelf. Barcode and QR scanning covers receiving, putaway, picking and shipping, with mobile screens for operators. Picks can be grouped into waves by zone, priority or carrier, and cycle counts run by location or ABC class.
Starting a production order in IRIS MES reserves its material in the warehouse, and finished goods are booked back. Receiving includes quality inspection with hold and release, so a blocked lot is not picked. Material shortages raise high-priority notifications, and AGV or AMR feedback can be connected to the same record.
IRIS calculates picking routes through the warehouse, groups orders into waves and forecasts reorder points from consumption. It also suggests slotting changes based on actual pick frequency, and a warehouse lead approves each one before anyone moves a pallet, with the approval kept in the log. Flow changes can be tested first in DBR77 DT. If your plant runs a separate WMS today, see How to Unify MES, WMS, QMS and CMMS Without Replacing Everything.
FAQ
Do we need a new WMS to supply the line in real time?
Not always. What matters is that order releases, consumption and quality holds reach the warehouse as they happen. If your WMS cannot receive them, that is the gap to close.
Should consumption be booked per piece or per pallet?
Per pallet, bin or count interval is usually enough. Booking at order close is what causes the delay.
Is a fixed route better than delivery on call?
For regular parts, yes. Keep calls for exceptions and track them, because a rising number of calls shows where the route or the minimums are wrong.
Conclusion
Start by measuring the minutes each line waits for material, by cause. If late picks and plan changes lead the list, connect order releases and resequencing to the WMS first; if holds lead, connect quality. Then let empty bins and counts create the replenishment tasks, and change routes or layout only where the waiting data points.
Sources
- Toyota, Toyota Production System
- Lean Enterprise Institute, Supermarket
- Lean Enterprise Institute, Kanban
- Lean Enterprise Institute, Milk Run