Discover in our video how to process logistics documents in ABBYY® Vantage.
Hello. Today I’d like to share with you how we process logistics documents. And typically when we get these sort of use cases, really what we see is that there’s a packet of information that’s been delivered to you that has logistic information and bill of lading documents, commercial invoices, packing slips, et cetera and our software has the ability to take that packet, split it, determine the document types, and then extract information from those documents. This is a very popular scenario in the logistics world.
Now, what we see in front of us here is a packet that I’ve received and it contained multiple documents. So this was one PDF with multiple documents in it. And our software was able to split it, and we were able to determine that in this case, the first series of pages is a bill of lading, and then I had a certificate of origin, there’s a commercial invoice on the document, and then lastly, a scale ticket. So in our case, what we have here is the ability to what we call assemble. So split up the packet, determine the document type, we call that classification, and then of course, based on that document type extract the information. So we’ve been able to determine that the first page, for example, is a bill of lading, and then the model that we call for this bill of lading is referred to here on the right. So this is the information that we’ve extracted on the software.
The cool part about this technology, especially in transportation, is that we have the ability to machine learn different document types. So as we see different bill of ladings or as we see different commercial invoices, the software can learn those documents and also it can learn what to extract off of those documents. So not only does it know that a new document is present and how to classify it, but it also can determine what to extract on that document, even though it’s never seen it before. So machine learning in logistics is a very, very important aspect of automating these sort of processes.
Now that you understand the concept, let me show you a little bit behind the scenes on a typical setup for a ABBYY Vantage Process Skill. What you’ll see here is that we have always an input and an output. Every single process skill will always have that, but right here in the middle is really where the critical part of this process skill comes into play. You see, I have an assemble step. The assemble step gives me the ability to classify and break apart that packet. So although I may have one PDF with all of these details in it, or one email with all these details in it, we can determine when to split it. Now that we know how to split it, we will classify it. So we’ll determine the different document types within the packet, and then of course, based on those document types, we will extract it. Then lastly, if and only if needed, we would send that to a review queue. So this is a very common flow, especially in more complicated scenarios where we just need a ton of document types in one given packet.
Now, before I wrap this up, I wanted to kind of share also the different document types that we support. We support a tremendous amount of document types in the transportation and logistics world. You’ll see some of those listed here. If you don’t see some of the documents that you have, it is very likely that we have models already created for them. Also, our machine learning technology gives us the ability to adapt to new documents that we’ve never even seen before, very, very rapidly, and without the need to do templating or without the need to have consultants or highly technical staff to do that for you.
Definitely continue looking into your logistics use cases here with ABBYY Vantage. I think we’ve got a really cool piece of technology that can automate this for you.
[Music- “Engineered to Perfection” performed by Peter Nickalls, used under license from Shutterstock.
ABBYY, ABBYY Vantage, Vantage, and the ABBYY Logo are all registered trademarks of ABBYY Software Ltd.]