Bridging the Gap between Paper and Data

May 19
08:05

2005

Tokairo

Tokairo

  • Share this article on Facebook
  • Share this article on Twitter
  • Share this article on Linkedin

The cornerstone of successful automated office systems is the ability to convert printed information into electronic data. Document processing applications need to capture and index data accurately and efficiently to bridge that gap.

mediaimage

This capability can be enhanced with the integration of Optical Character Recognition (OCR) software. By this means,Bridging the Gap between Paper and Data Articles extracted data is used to index and save the document into the document management application.

This intelligent document recognition and classification is central to Tokairo’s new contract with a major UK supplier of dairy products. The delivery and receipt procedure works like this:

  • When an order is received from a customer a despatch note is issued listing product description, quantities, weight and code
  • On delivery the note is signed by the customer along with the customer’s Goods Received Note (GRN) which details the goods actually delivered

Problems arise if goods are damaged, lost or delivered piecemeal, as discrepancies can occur between the despatch note and the customer’s GRN.

The system has been set up so that these documents can be automatically read, matching delivery line items with corresponding items from the company’s despatch notes.

After every delivery, despatch notes, GRNs and any other delivery/receipt documents are scanned by the delivery driver at any one of the company’s 20 depots in the UK.

Coping with different document styles and formats

The company’s system is configured to recognise every GRN’s header and footer, so it can read each item’s line code and quantity. This process is made complicated because each customer uses a different GRN style, format and layout. Also, there can be multiple GRNs for one despatch note – typically when a single delivery goes to several different customer sites.

However, once the OCR software has been correctly configured to read the lines in the different styles of document, accurate reading of forms is high – ranging from 85% to 90%. A threshold can be set by users so that any document that falls below a certain level (e.g. 80%) is entered manually. So any document that can be read less than 80% accurately is flagged up for manual indexing. Anything above this value is saved automatically.

The system automatically consolidates this process by matching the different documents and line items in the original order, and stores all documents in the appropriate order folder.

Automatic invoice amendment

Any discrepancy is automatically flagged up and investigated. This generates an amendment to the sales invoice issued to the customer, which in turn reduces the company’s debtor days by ensuring that the sales invoice matches the customer’s GRN.

Should there be any discrepancies or delays in receiving payment, staff have online access to all delivery information stored in the order folders.

Integrating intelligent document recognition and classification technology provides significant customer benefits:

  • Reduced administrative overheads
  • Improved business visibility
  • Reduced invoicing errors and attendant disputes and delays
  • Increased cashflow by minimising debtor days
  • Improved customer service
  • Enhanced corporate image