White Furniture? is one of the leading suppliers of medical furniture in Russia with customers from all over the country.

Most items at their warehouse are sets of separate packages. E.g., a bed consists of three packages: a bedstead, head- and footboards, and a mattress.
Packaging of different versions of the same product doesn’t differ much in appearance.
On top of that, components of a set can be sold separately, not as a part of that set

The problems that arise are evident:

• unacceptably low order picking accuracy (packages from different sets get mixed up);
• hence is mixing up in dispatched consignments (which is a particular fun, if the shipment goes to a destination 5000km away);
• hence is customers being unhappy, lots of pain in the ass, and overheads from dealing with this pain;
• the mirrored mixing up of goods at own warehouse leads to physical piles of incomplete products, gaps in stock and losses in the form of dead working capital?.

The listed problems are obvious to anyone who deals with warehousing in real life.
Those who have had priceless hands-on experience dealing with a warehouse filled up with composite goods and without a proper IT support, may add some vivid details, like papers, duct-taped to boxes, with a 42 pt size text: "a piece of product 1199365. Not to confuse with product 119389. $300 fine."

Fortunately, back in the day White Furniture was lucky enough to obtain? a proper IT support with dia$par.

This is how things turned out.

To solve the problem of organizing a reliable and ergonomic processing (and keeping record) of composite goods, we used the mobile data terminal, which, among other things, implements bin locations storing, and keeps track of all operations on individual basis?.

For each product, all of its components are listed. Each component means a separate package in the warehouse.
?dia$par keeps double track of the stock — by product as a whole and by its components.

Based on documents for expected goods receipt, stickers for product labelling are printed out.
If it’s a composite product, stickers for each of its components are printed out.

The sticker indicates the component’s name, what product it is part of, and a total number of components for this item.

A component that comes to the warehouse not as part of a set, is indicated by a different sticker configuration and information on it.
By default, such component will be shipped off separately as a separate product.
A technological procedure of "completing a composite product" is provided for a virtual assembly of such components into complete sets. During such procedure the future components of a single set will be re-marked with newly printed out stickers.
The system supports a reverse operation of breaking down sets into single components, whereby the components are also re-marked.

When products arrive at the warehouse, an internal warehouse acceptance form is printed out. For composite goods, it indicates their components.
It also indicates locations of bins, where these articles are currently stored.

By scanning bin and product barcodes??, warehouse employees tell the system in which of the bins the product is placed and in which quantity.

That said, composite goods components are processed in exactly the same way as any other product — the logic of products being complete is kept by dia$par on its own in the background mode.

In a similar way, for order picking, an internal warehouse delivery note is printed out.

The PDA of an employee picking an order indicates components and bins locations.

In case there are any discrepancies between the system data and actual bin contents, the warehouse employee uses his PDA functionality to perform inventory count on the fly, which takes just a few seconds, without having to access a desktop computer. Thus, system data becomes up-to-date again, while dia$par informs the responsible personnel about this unpleasant event and the need to start investigation.

dia$par logs data for every scan that employees make (its author, PDA number, exact time, etc.). Individual track of transactions is kept according to employees" IDs, printed as barcodes on their badges.

While solving the composite goods problem, we became aware of the situation that the system has no valid data on actual product dimensions — that is the dimensions of packaging.
The system contained only dimensions of products proper, provided by vendors, and these were used to calculate truck load factor as part of automatic route calculation?.
For obvious reasons, dimensions of furniture-like product packaging have a particularly great difference from dimensions of products proper. In the end, we introduced measuring of packaging as a step of goods acceptance business process.
Foolproofing: items without filled-in packaging dimensions are highlighted in acceptance documents and on PDA screens; the system doesn’t allow such products being placed into bins.
Foolproofing 2: direct input of dimensions into product properties is forbidden. A special ruler with barcodes on top of numbers has to be used for measuring. So in order to enter new product packaging dimensions into the system, it’s necessary to perform the following:

• scan the sticker barcode with the employee’s PDA;
• for composite goods — the component’s barcode has to be scanned as well;
• place the barcoded ruler next to the packaging;
• scan the spot on the ruler, where the packaging ends.

Practical results of dia$par innovations:

• mixing up goods during delivery is now a real force-majeure — while it used to be a humdrum of life before. And it takes only a minute to learn who’s responsible for it and to find out the nitty gritty.
  Investigating delivery screw-ups with customers underwent similar changes, upgrading from the former "daily-pain-in-the-ass" status.
• quantity of mixed up goods in own warehouse decreased by more than 30 times.
• the complete physical inventory count process is now significantly simpler.
  Actually, it’s better to say that for the first time in the company’s history this complete physical inventory count became in fact possible, while its results — valid, interpretable and practically applicable.
  Inventory count results are automatically recorded by the system immediately in the course of the counting process, which in its turn can be described as "scan as you go".
• Warehouse personnel work became standardized? to a high degree.
  Now, if two or three "warehouse gurus" are laid off, it won’t put the business into a "deliveries almost on hold" situation, whereas training of newly hired troopers to being field-level eligible takes only a few hours.