Warehouse management: on the "waves" of operations

When confronted with automation, one often hears about “wave assembly” and “wave orders”. For a start, a little introduction on how the term "wave" appeared. Imagine a graph where the X axis is time and the Y axis is a certain indicator. We wait until the value of the index along the Y axis increases to the specified maximum value, and we begin to perform actions leading to a decrease in the value. Then, “falling” to the specified minimum value, we again wait for accumulation. Thus, we have a graph in the form of a kind of "waves." I will give a simple example, when the use of such “waves” may be beneficial for us when assembling orders:

1) The warehouse receives a lot of orders, each of which in its volume is about 1/10 of the volume of a whole pallet with a height of 1.8 meters;
2) An employee can take one pallet for a set and collect 10 orders at once, bringing it to a height of 1.8 meters, which will not prevent him from moving it;
3) We wait until the warehouse receives 10 orders, after which we send the employee to recruit them all in one pass.

One of the main tasks in order picking is to reduce the picker's route. For comparison: the selection of the 1st box from a cell takes 3 seconds, and moving between cells can take 30 seconds, depending on the structure of the dialing zone. As a result, if we issue one order to the order picker, he will run to collect it, pick up 1/10 of the volume that he can physically move, and come back, following the rather big route, as this order may include nomenclature items very far apart. If we put together 10 orders at once, then the employee, having followed a similar route, will be able to collect 10 times more.

However, in addition to the optimal value at the level of the same volume for picking, we have another value that we have to take into account: time. If we wait for a very long time until we have the optimal volume, we may lose precious time, and perhaps - if we have, for example, a warehouse-store - a customer who does not want to wait too long. That is why the "waves" may not be the same height, but the essence does not change.
')
This concept is very similar to the concept of “batching” (batching), when we try to combine disparate tasks into groups in such a way as to minimize the labor costs for their implementation. For example, what's the point of sending a stacker to a nearby room to perform a single task if it has enough of them in the current location? Of course, specialists experienced in these issues will probably notice that each operation has its own priority, and they will be right in their own way, but the question of prioritization contains so many nuances that I would prefer to devote a separate article to it, and now we will leave it behind.

With regard to the mentioned stacker, moving it to the next room alone, as well as returning back after completing the task, can be more labor-intensive than performing 2-3 tasks here and now.

Realizing that the "waves" can be used in relation to almost any parameter, consider the most commonly used options:

1) "Waves of orders" - the accumulation of a certain volume of orders to be sent to work in order to optimize the loading of performers and reduce the route of picking;
2) “Waves of Displacement” - waiting for the achievement of a predetermined value in the formation of tasks for the execution of movements of pallets inside the warehouse in order to reduce the time of operation of heavy machinery;
3) “Wave Assembly” is a term that most often means waiting for so many orders that would allow to take an entire pallet with a certain product to a separate section and sort it in full (or close to full) volume for each order.

I’ll focus separately on option 3. It is great for those warehouses that work with a small number of nomenclature positions, but at the same time receive a large number of orders. In this case, we can not go around the entire warehouse, picking up orders, and carry out the distribution of the specified nomenclature positions in the pool of orders that are being processed. Suppose you have 100 orders, each of which contains the 1st box of item X. One pallet (pallet with the item) contains, say, 100 boxes. Thus, we can take a pallet, transport it along order picking sites, and distribute it over them, or put it in the center of a high-intensity distribution site, and allow employees to take the quantity required for each order at the repacking or control stage, reducing the labor costs for equipment account significant route reduction.

I would also like to note that orders can have a different structure, and it is classic to divide a given quantity for each position into three components:

1) The fact that you can take out whole pallets;
2) The fact that you can pack in whole packages;
3) What will be collected pieces (minimum units).

A well-designed WMS allows you to work with each such component separately, and use different grouping algorithms for “waves” and “batches”, associate tare calculations, sequences of execution, combination and priority, and much more with such groups.

Finally, I want to note that the “waves” are about the same tool as the notorious “ABC-analysis”, which everyone knows, but not everyone is able to competently use.