Millions of packages and parcels move across international borders every day — and everyone involved wants to know what’s being transferred and by whom. Do these parcels contain dangerous, illegal or prohibited goods? What risks are associated with the shipper? Can customs duties and taxes be properly assessed and collected? How can customs clearance be expedited?
The answers to these questions lie, of course, on the customs label that the shipper completes before dropping the item into the delivery network. The information from the shipper can be captured online very effectively. Various programs are available (such as the global Data Sharing Agreement from International Postal Corporation) to assist in the transfer of the data. Where information is not provided online, optical character reader (OCR) technology may substitute for manual entry. And, as the effectiveness of these data-capture systems improves, data becomes more complete and accurate.
Today’s realities
But here’s the unfortunate truth: much of the customs data is never captured. As parcel volumes and the number of shippers continue to increase, postal and customs authorities simply can’t keep pace with all the customs labels that weren’t populated through an online system or automatically read through OCR. The cost of manual entry is prohibitive, so the packages cross the border unsecured and without collection of fees.
International standardization of customs forms would go a long way toward addressing this challenge. With such standards, forms-reading technology could be deployed effectively to capture the information from the labels. We know, however, that each country is free to individually design the forms, even if the content is standard. Conventional data capture from forms relies on forms templates and OCR to locate and classify the information on the form. With no standardization, the usual approach to forms-reading can fall short in the face of a myriad of layouts, languages, sizes and presentations.
So, what is a postal authority supposed to do? As any organization that uses parcel automation knows, simply locating the address label among all the noise on a parcel, then reading address information, can be difficult. And the difficulties multiply with reflection, wrinkled pouches, straps and odd shapes. Reading customs forms can be even more problematic and requires a solution that goes beyond the conventional template-based methods.
Tomorrow’s postal services
The solution to this problem must not only address the current situation but also enable an agile and adaptive post of the future. Capturing customs data from images lifted from automated parcel sorting machines, or from hand-addressed packages dropped at the counter, requires highly advanced recognition technology. Effective capture necessitates innovative techniques that can obviate the need for label extraction and classification while operating irrespective of translations, transformations and rotations.
Unlike the traditional approach to forms-reading, advanced systems can apply image processing, machine learning and complex algorithms to overcome the challenges inherent in reading parcel and package images. And this technology must be available on high-speed sorting machines as well as on mobile devices at the counter.
Did you know it’s already possible to obtain complete, reliable data for transfer to the destination country, the receiving country and the customs organizations on either end of the journey? Problems in capturing customs data are readily solvable with access to recognition capabilities that defy the norm. In our conversations with customers, we’re seeing huge interest in this area. These capabilities are particularly powerful when they can be easily integrated and modified alongside existing or new sorting machines.
That’s how digital solutions that solve today’s challenges can enable essential agility that will define the post of the future.
