OCR vs RPA vs AI (Shipamax): Which approach delivers the best approach for logistics back office automation?

OCR vs RPA vs Shipamax

When speaking with freight forwarders, we're often asked "Is Shipamax different to OCR and RPA technology?"

The short answer is yes.

• Shipamax does not compete with either OCR or RPA
• Shipamax makes use of OCR to extract data from images
• Shipamax uses a proprietary 'contextual understanding layer' to add meaning to text
• Shipamax compliments RPA, working as a first step to push unstructured data into any RPA technology for enhanced back office automation and digital transformation

In this article, we'll go into detail on the technology behind Shipamax, how OCR and RPA work with Shipamax and the benefits for freight forwarders when adopting a specialist logistics automation technology.

What is Shipamax?

Shipamax is a plug and play toolkit for back office automation in logistics. Our core product connects to any email inbox or unstructured data source and automatically extracts data from emails and attachments in real-time - outputting a clean, structured feed via API. We turn unstructured documents into data, specifically for the logistics industry.

How does Shipamax work?

There are two parts to the document parsing process:

Step 1: Classify

We use machine learning models to classify logistics and supply chain document types, e.g. bills of lading, commercial invoices and accounts payable invoices.

Step 2: Extract & Understand

After classifying a document, Shipamax extracts data from the document. If it's image based, we'll use OCR to extract that data. After extracting the text, our machine learning models understands the context of the document. This enables the previously unstructured data to be structured and mapped to our document schema, for example, a Master Bill of Lading schema. This schema can be pushed into your ERP or TMS such as CargoWise.

What is OCR and how does it work?

OCR or optical character recognition, is a technology that allows text to be extracted from image based documents, such as PDFs or scanned invoices. The data is converted into digital data or editable text.

For example, imagine you've been sent a paper document, like a scanned bill of lading or accounts payable invoice. OCR technology extracts the data so you can copy & paste that into a system.

OCR technology has its limitations. Whilst it's excellent for data extraction, it doesn't understand context. 'Smart OCR' solutions sound great, but often require you to create templates and rules to interpret data and add meaning to that text. With so many document variations for freight forwarders to handle, these rules can break. Maintaining these can take a huge amount of time and resource.

What is RPA and how does it work?

RPA, or robotic process automation, allows you to automate processes that are rule-driven & performed on a computer. If you instruct an RPA on what you need it to do, it will be able to run every second of every day performing that task. However, if a task is not rule-based, RPA may not be the right tool for the job.

In the context of extracting data from documents, this is where RPA differs from a machine learning based platform like Shipamax. When humans read documents, there is judgement involved - this is not a fully rule-based process. So just like OCR, RPA can struggle to deal with unstructured data sources, like invoices or bills of ladings.

What different data types can each technology be used for?

• OCR is a great tool for translating images to text.
• RPA is used to automate rule-based tasks performed on a computer. Usually structured data already exists in the system.
• Shipamax is used to take unstructured data and convert it into machine-readable data. Data can be provided in a number of different formats - image, email, HTML, Excel or Word.

Shipamax and OCR

Shipamax has an OCR component built into its platform to assist with data extraction from bills of lading, accounts payable invoices, commercial invoices and packing lists.

Shipamax uses AI, or what we like to call, Anti-Template Technology, which means, no more creating templates or rules on top of OCR. When extracting text from a document, our machine learning understands what that text means, and accurately maps it to a clean data schema, which can be fed into your ERP system.

When paired with an AI-powered solution like Shipamax, OCR becomes a much more powerful technology. There is no need to build templates and create additional rules to understand new documents. The OCR within Shipamax will carry out the data extraction, while the machine learning engine will learn from each new document it reads, and structure the previously unstructured data. If you're interested in learning more about OCR, AI and data extraction from account payable invoices, here we discuss some of the limitations of OCR as an AP invoice reader.

Shipamax and RPA

The biggest challenge in extracting data from logistics documents is dealing with the multitude of invoice formats suppliers and manufacturers use across the industry. This is where Shipamax complements RPA. Shipamax extracts unstructured data from any type of invoice or email, and then standardises that data into a structured format. Once Shipamax has automated the data entry into your ERP, an RPA can then be used to provide rule-based tasks off the back of this. Once you've combined your RPA with Shipamax's machine learning engine, you'll be able to automate workflows for both structured and unstructured data.

So how do Shipamax, RPA and OCR work together?

So what are the benefits of adopting a specialist document automation tool like Shipamax?

There are a number of different approaches you can take when implementing a back office document automation process. When weighing up options, there are three key areas to take into consideration:

• Accuracy
• Integration
• Workflow

Option 1: Traditional OCR plus internal development work

The first option many companies consider is purchasing a stand alone or traditional OCR. The issue here is that OCR is only a tiny component of a logistics back office automation solution. Purchasing an OCR means you still have to go and build the rest of the solution in house. This will take a significant financial and workforce investment.

Option 2: Shipamax

Shipamax specialises in providing automated data extraction to freight forwarders and the wider logistics industry. The solution is 'plug and play' - three key benefits of this are:

1. No set up or maintenance - Our proprietary, anti-template machine learning models have been trained across millions of logistics documents. We continuously train these models. You'll never have to draw a template again.
2. Resource-lite roll out. We've all been there - the nightmare edge cases of integrations that drag out for months. Our out-of-the box integrations with popular ERP systems such as CargoWise mean minimal work for your internal tech team. We've done it, so you don't have to.
3. Seamless workflow - When rolling out this technology within your organisation, you'll need to consider workflow. What happens to those emails that have been automatically processed? What happens if something gets rejected by CargoWise? We've carefully crafted workflow solutions to sit natively within your current systems & processes.

The most suitable approach for your business will ultimately come down to a number of factors, not limited to costs and workforce. If you are considering investing in a specialist data extraction software for logistics document, why not join one of our 15 minute live demos, where you can find out more about Shipamax.

We'd love to talk

Shipamax helps free logistics companies from manual admin using plug and play automation. If you're a freight forwarder and you'd like a free consultation to run through your options, we would love to hear from you.