How Systech Vision drives the aggregation excellence of Pack-By-Layer

Jun 21, 2024
By Tarek Ismail

In part 2 of the vision series blogs, we’ll be talking about Pack-By-Layer.

Pack-By-Layer is an aggregation solution where all items packed into a layer of a package are read by the vision system with one high resolution image. There’s lots of advantages to this in manual processes as it relieves operators from having to manually scan each item prior to packing, which is time-consuming and prone to error.

Having an overhead camera that reads as you pack, and you can look at the screen to check the process sounds like a much more reliable and user-friendly operation. This also works quite well with automated processes as it decreases the risks involved, and there’s some history there.

Before camera technology caught up, from a resolution perspective, all aggregation was done in an infeed sort of way: reading one item at a time, or maybe a group of items that make up a row of a layer – not one whole layer, and through signal exchange, the serialization engine can track those items as they are formed into rows, and then layers and then finally placed or pushed into the case – which comes with the assumption that whatever was read and accepted at the infeed will make it into the case. This is generally fine and works well in the majority of situations…and even if it wasn’t, there were no other choices because reading more items at a time was not possible due to camera resolution limitations.

This infeed method comes with some risks due to the many “what ifs” that the real world can throw into the case packing process:

  • What if two cartons are not separated and the sensor doesn’t catch one of them?
  • What if a whole layer is ready to be picked but something stops the machine or crashes the pick and place mechanism?
  • What if operators don’t clear the machine properly leaving items in between phases of the operation?

These are just some of the things that could throw off the aggregation process that relies on infeed reading. Many of these risks could be avoided if the reading of what’s getting packed happens when the layer is fully built. The hard part is the vision side – making sure this is doable now that we have these fancy cameras.

The solution from a Systech vision perspective has gone through iterations of improvements, each taking lessons from what has come before it. It was introduced as a new feature within our existing data matrix tool to enable it to read in grid mode. From the onset, the direction was to have the grid details (number of rows x columns) come down from the Systech level 3 system as part of the product setup. As more projects were deployed, the requirements and challenges for such a vision inspection became much clearer which led to the creation of a brand-new tool that was specifically made for reading a grid of data matrix codes with features that enable it to function in this environment.



This new tool, tailor-made for Pack-By-Layer, was built with a different kind of training functionality that made the inspection of each layer unique when it came to the specifics of how the tool finds and reads the codes. In manual, or top-load automated case packing operations, each layer is expected to be at a different distance from the camera, which means each layer will have codes that are different in size from the other layers, and the location of the codes in the image will also be different as each new layer is closer and closer to the camera. Having the tool built with this as its theory of operation enables reading of smaller codes, faster operation as the tool knows better where to expect the codes in each layer, and higher read rates because it also knows the size of modules in the codes for each layer.


This is all great, but it wouldn’t be possible without having the ability to automatically refocus the lens when these new layers are introduced – so, innovation on the hardware side was also needed to enable the great ideas the vision software ran with. As a new layer is placed, the software sends focal distance instructions to a servo motor attached to the camera that adjusts the focus in the lens to produce a sharp picture that the vision tool uses to read the codes. It’s beautiful, elegant, and works well too, thanks mainly to the innovation in vision. This is not theoretical, there are sites that rely on it in their daily high-volume operations, packing anywhere from a few to hundreds of cartons per layer.


To learn more about this innovative solution Systech offers comprehensive, multi-faceted training programs—tailored to your knowledge level and goals—to ensure every customer is fully trained on our powerful vision, serialization, traceability and authentication solutions. View our current course offerings here.


Related posts