Glossary

Order picking – methods, systems and key figures

Order picking is the assembly of articles from a range on the basis of an order. Basic principles, methods, guidance techniques and key figures explained.

Order picking is the assembly of articles from an overall range on the basis of an order. In most warehouses it is the most labour-intensive process, and it largely determines lead time, delivery quality and logistics cost.

Two basic principles

Person to goods: the picker moves to the storage location. This is the classic method in shelving and pallet warehouses — flexible and low in investment, but travel-intensive. The travel share frequently accounts for more than half of the picking time.

Goods to person: the articles are brought to a workstation automatically, for instance out of an automated small parts warehouse or a shuttle system. Travel time disappears and output per workstation rises; investment and dependence on the system are higher in return.

Methods and guidance techniques

Order processing can be organised in a single stage (order by order) or in several stages (article by article with subsequent consolidation). Common forms are single-order picking, multi-order picking and zone picking with handover.

For guidance and confirmation, the field uses pick lists, mobile data terminals, pick-by-light, pick-by-voice and increasingly pick-by-vision. The choice follows from pick density, the cost of an error and the structure of the range.

Key figures

Picking is assessed through picking performance (lines or picks per hour), the error rate (often stated as errors per 1,000 picks), the travel share and order lead time. These figures are comparable only in the context of the order structure — few lines per order produce different values than large orders do.

Finding a system according to VDI 3590

There is an ordered procedure for the path from data to a system decision. VDI 3590 Part 2 describes it in three steps: gather the basis, build variants from modular kits, eliminate variants against criteria.

Base data

The layout, size, technology and processes of a picking system are determined largely by its base data. The guideline groups them in four:

  • Article structure data — physical data such as dimensions, weight, shape and hazardous-goods marking, stackability, surface properties and sensitivity to pressure, together with the assignment to article groups and load carriers.
  • Range data — number of articles, ABC structure by sales or volume, number and extent of article and product groups.
  • Sales structure data — static values such as units per load carrier, and dynamic ones such as total stock, minimum stock, sales per unit of time and access frequency.
  • Order structure data — the number and composition of orders, from which lines per order and order peaks follow.

Two instructions in the guideline are regularly passed over in practice. First, average and maximum data are always to be considered — a design based on mean values collapses at the peak. Second, data collection is not a one-off exercise but is to be repeated across planning and operation so that changes can be answered.

Three modular kits

Solution variants arise from three morphological kits that the guideline keeps separate:

  • Organisation system — single-zone or multi-zone. Zones are organisational or technical units with similar properties, for example fast-mover zones, reach zones or areas of identical racking. The process organisation belongs here as well, such as collecting a pick list split across several units either one after another or simultaneously.
  • Information system — order capture is manual, semi-automatic or automatic. What governs the choice is the number of transactions per unit of time and the required total lead time of an order. Confirmation belongs here too: per unit removed, per line, or jointly for all lines.
  • Material flow system — the movements of the system elements directly involved in removal, together with the movements upstream and downstream of it.
Base data article structure data range data sales structure data order structure data average and maximum Organisation system Information system Material flow system three modular kits that variants are built from Decision criteria qualitative: flexibility, universality, modularity, availability quantitative: from the base data System decision The criteria eliminate variants that violate a condition early — what is examined is not every conceivable combination, but the admissible one.
Finding a picking system — own illustration after VDI 3590 Part 2, section 2.

Decision criteria

Alongside cost, the guideline draws on qualitative and quantitative criteria. The quantitative ones come from the base data. The qualitative ones it defines in its own terms:

  • Flexibility — the short-term ability to adapt to fluctuations in the required picking performance.
  • Universality — versatility in application, related either to the variety of the range or to possible future uses.
  • Modularity — systems made of several independently functioning subsystems that can be extended piece by piece.
  • Availability — of particular weight in automated and mechanised systems, because there the failure of one element determines the throughput of the whole.

The purpose of these criteria is to eliminate as early as possible the variants that violate a condition. For the comparison between the current and the target system, the guideline additionally names cost figures such as cost per pallet location or per article provided.

Where picking sits in the warehouse

Picking is not an isolated activity but one function among several. VDI 3629 "Organisational basic functions in the warehouse" (edition 2005-03) — written for planners and operators alike, and related to customary unit-load warehouses such as block, rack and high-bay warehouses — lists twelve of them:

Twelve basic functions — picking sits between put-away and retrieval 1Goods receipt2Distribution to storage areas3Identification point (I-point)4Put-away5Warehouse administration6Retrieval7Relocation8Checking (K-point)9Distribution to dispatch zones10Goods issue11Stocktaking12Reporting The functions may also be performed in upstream or downstream areas.
Organisational basic functions in the warehouse — own illustration after VDI 3629, section 3.

Two points of this systematics are regularly underestimated in projects. First the identification point (I-point) and the check point (K-point): they are functions in their own right with their own space, staff and time requirement, not side activities of goods receipt. Second the guideline's note that the basic functions described may also be performed in upstream or downstream areas — whoever looks for them in the warehouse and fails to find them there has not abolished them, only moved them.

VDI 3629 also names what warehouse organisation aligns itself to: article variety, turnover frequency, turnover quantities and the technical equipment. These four precede every decision about a picking method.

Where it sits in planning

Sizing the picking operation is part of warehouse planning: methods, zones and technology follow from article and order structure data — ABC distribution, lines per order, order peaks. In practice it is decided together with the choice of storage technology and the warehouse management system. The step-by-step procedure is set out in our guide to selecting and sizing a picking system.

Sources

VDI 3590 Part 2 "Order picking systems — finding a system". VDI 3629 "Organisational basic functions in the warehouse", edition 2005-03, VDI Handbook Material Flow and Materials Handling Volume 8; that edition was confirmed against the VDI guideline list on 5 September 2026.

Order picking in practice: Our services overview brings together the relevant planning and consulting approaches.

FAQ

Frequently asked questions

What data does the planning of a picking system need?

VDI 3590 Part 2 names four groups of base data: article structure data, range data, sales structure data and order structure data. What matters is that average and maximum values are collected for each — a design based on mean values alone will not carry the order peak.

How do person-to-goods and goods-to-person differ?

With person to goods, the picker moves to the storage location; the method is flexible and low in investment, but the travel share frequently exceeds half of the picking time. With goods to person, the technology brings the articles to a workstation: travel time disappears and output per workstation rises, while investment and dependence on the system are higher.

Which criteria govern the choice of a picking system?

Alongside cost, VDI 3590 Part 2 draws on four qualitative criteria: flexibility as short-term adaptation to fluctuating performance, universality as versatility towards the range and future use, modularity as extendability piece by piece, and availability. The quantitative criteria come from the base data. Their purpose is to eliminate inadmissible variants early.