Contents
Orders are going out late. The 3 pm carrier pickup leaves with half a dozen orders still on the pick list, overtime has become normal, and the first answer is always the same: hire another picker. Before you do, follow one of your pickers for an hour. Most of that hour is not spent picking. It is spent walking.
Part 1 of this series won back space in the example distributor's pallet-rack zone and moved the fast movers to the front. This part follows the pickers through the same warehouse. It compares the four common ways of picking orders and the two simple routes through the aisles, and shows how many picker-hours each one takes for the same day's work.
Where a picker's day goes
A pick trip has four parts: getting the list and the cart, walking between locations, picking the case off the pallet, and sorting or consolidating the result for packing. The order-picking literature has long reported that travel is the largest of these, often about half of a picker's time or more, which is why most of the ways to pick faster are ways to walk less.
The example picks 899 cases a day from the rack zone: 300 orders of about 3 lines each. The times assumed throughout are:
| Assumption | Value |
|---|---|
| Walking speed with a cart | 0.9 m/s |
| Pick one case off a pallet | 20 s |
| Put a line into the right order's tote (batch, wave) | 4 s |
| Set up a trip: list, labels, cart | 60 s |
| Set up a zone trip | 30 s |
| Bring an order's totes together at packing (zone, wave) | 15 s |
| Productive picking hours per picker a day | 7 h |
With single-order picking in the as-is layout, each order costs about 192 m of walking. The day takes 27.8 picker-hours, and 64% of that is walking. That is 4 pickers' worth of work for 899 cases.
The four pick methods
| Method | How it works | Suits | Watch out for |
|---|---|---|---|
| Single-order | One picker takes one order from start to finish, then comes back for the next | Few orders a day, large orders, urgent one-offs | Walks the whole warehouse for every order. Picker-hours rise in step with orders |
| Batch | One trip picks several orders at once onto a cart of totes, one tote per order, and each line goes straight into its order's tote | Many small orders: e-commerce, parts and service orders | A put-to-the-wrong-tote error. Needs a scan or a clear tote label at each put |
| Zone | Each picker owns one area and picks only the lines in it. An order's totes meet at packing | Larger buildings, several storage types, several pickers a shift | Uneven zones leave one picker swamped and another idle. Consolidation at packing |
| Wave | Orders are released in timed waves set by the carrier pickups. Each zone picker sweeps their zone once per wave | Fixed carrier cut-offs, steady daily volume | Needs a WMS or a disciplined release routine. A late wave misses its truck |
The methods combine. Wave picking, as modelled here, is zone picking with orders released in 3 waves a day, timed to the pickups. Each zone picker carries up to 40 lines a trip. Batch and zone trips here hold 8 orders.
Routing: return or S-shape
The route matters as much as the method. Two simple rules cover most small warehouses, and both can be printed on a pick list as a sorted location sequence.
In the example the answer flips with the layout. In the as-is warehouse, fast movers are scattered along the aisles, so the S-shape route is a little shorter for single-order picking (186 m against 192 m an order). After Part 1's re-slot, the busy pallets sit in the front bays, and the return route wins clearly: 92 m against 136 m. Slotting and routing have to be decided together.
Figure 2 puts pickers into the Part 1 warehouse. Each coloured trail is one trip, and each glowing dot a case pick. Choose a method, or Compare all to see the four side by side, switch between the as-is layout and the layout after Part 1's fixes, and move the slider to change the day's orders. The numbers come from the same model as the tables.
This browser cannot show the 3D model. The controls below still update the numbers.
- Single-order
- Batch
- Zone
- Wave
- Case pick
The result
The table gives each method with its better route, at 300 orders a day, in both layouts.
| Layout | Method | Route | Walk per order | Lines per hour | Picker-hours | Pickers | Walking |
|---|---|---|---|---|---|---|---|
| As-is layout | Single-order | S-shape | 186 m | 33 | 27.2 | 4 | 63% |
| Batch | S-shape | 46 m | 82 | 10.9 | 2 | 39% | |
| Zone | Return | 50 m | 72 | 12.4 | 2 | 37% | |
| Wave | Return | 10 m | 107 | 8.4 | 2 | 11% | |
| After Part 1 | Single-order | Return | 92 m | 49 | 18.5 | 3 | 46% |
| Batch | Return | 26 m | 99 | 9.1 | 2 | 27% | |
| Zone | Return | 17 m | 98 | 9.1 | 2 | 17% | |
| Wave | Return | 5 m | 113 | 8.0 | 2 | 6% |
Three things stand out.
- Slotting comes first. Part 1's re-slot alone, with no change of method, cuts single-order picking from 27.8 to 18.5 picker-hours a day.
- Batching halves it again. Picking 8 orders a trip brings the walk down to about 26 m an order and the day to 9.1 picker-hours. Two pickers can now do what took 4.
- Zone and wave win on walking, not always on hours. Wave picking walks only about 5 m an order, but it pays for consolidation at packing. At 300 orders it saves about 1.1 picker-hours a day over batch picking, and it needs the zones staffed and balanced.
Volume changes the choice. At 150 orders a day, batch picking takes 4.5 picker-hours and fits one picker. At 600 orders, single-order picking would need 36.5 hours, or 6 pickers. Batch takes 17.5 hours and wave 15.4, because each wave's trips get denser as the orders grow.
Choosing a method
| Your situation | Start with | Then |
|---|---|---|
| Under about 100 orders a day, one picker | Single-order with a sorted pick list and good slotting | Batch two to four orders a trip when the orders are small |
| Many small orders, one to three pickers | Batch picking with a tote per order on the cart | Add scan-to-tote to stop put errors (Part 5) |
| Several pickers in the same aisles | Zone picking, with zones sized by picks, not by floor area | Re-balance the zones every quarter |
| Fixed carrier pickups that keep being missed | Wave release timed to each pickup, with batch or zone picking inside each wave | A WMS to release and track waves (Part 7) |
Measures to track
| Measure | What it shows |
|---|---|
| Lines picked per picker-hour | The headline productivity number. Track it by method and by person |
| Walk per order or per line | Slotting and routing at work. A handheld's step count or a few timed trips are enough to start |
| Orders shipped by the carrier cut-off | The number customers feel |
| Picks from the front bays and the floor or first beam | Whether slotting is drifting (Part 1) |
| Pick errors per 1,000 lines | Whether speed is being bought with mistakes |
Common mistakes
- Hiring before measuring. Time five trips. If walking is half the trip, the fix is the route and the method, not another person.
- Unsorted pick lists. A list in order-entry sequence sends the picker back and forth. Sort it by location along the route.
- Batching without a put check. Faster picking is lost if the wrong case lands in the wrong tote.
- Zones by floor area. Size zones by picks a day, or one picker carries the fast movers for everyone.
- Changing the method before the slotting. The best route and method depend on where the fast movers sit.
Key points
- Walking is the largest part of a picker's day. In the example it is 64% of single-order picking as-is.
- Re-slot first, then batch: together they take the example from 27.8 to 9.1 picker-hours a day.
- Choose the route with the layout. Return suits fast movers at the front; S-shape suits picks spread along the aisles.
- Zone and wave picking cut walking most and suit higher volumes, fixed cut-offs and several pickers.
- Track lines per picker-hour and orders shipped by the cut-off.
Earlier in this series: Part 1 won back space and re-slotted the fast movers. Next: Part 3 goes to the dock, where receiving, put-away and cross-docking decide how fast stock becomes pickable. Related: ABC inventory analysis for ranking SKUs by picks.
Sources
- De Koster, Le-Duc and Roodbergen (2007), Design and control of warehouse order picking: a literature review, European Journal of Operational Research (travel share of picking time, routing and batching methods)
- Ramp: Warehouse slotting guide (travel share of pick time)
The distributor, its orders and its times in this series are an illustrative example built from stated assumptions. Measure your own pick rates and walking distances before choosing a method.
