Artificial Intelligence will impact all sectors of the economy and will change jobs. The customs world will not escape this technological upheaval. The concept of Artificial Intelligence (AI) refers to the ability of a machine to learn and make sense by copying human behaviors, interpretations and actions.
Artificial intelligence is a major technological development that draws much of its inspiration from biological systems, giving computers near-human abilities to reason and learn.
The processing of billions of commercial and logistical data with AI will make it possible to extract useful information for the customs professional in a few fractions of a second. From this point of view, AI is way ahead of man. It will be able to drastically reduce the complexities and slowness of customs.
A Deep Learning algorithm trained to reckon customs nomenclature for millions of products to achieve a reliability rate of over 97% for tariff classification. This is a much better performance than human. Artificial Intelligence makes it easy to move from manual to automated commodity classification which is faster and more reliable.
Fast growth of e-commerce has to face customs regulatory constraints, which threaten to slow down the fluidity of international delivery. Platforms and marketplaces like Amazon are eager to develop Artificial Intelligence, including customs, to optimize their logistics flow.
Briefly Artificial Intelligence will deeply change practices and transform the world of customs. This is a good opportunity to update the customs sector.
Artificial Intelligence as an essential tool in the customs world
Our point is not for Artificial Intelligence to replace human being. AI will not take the place of the customs worker, instead It will serve as a vital tool for customs professionals and increase their productivity.
AI can be used for several domains and different customs jobs. It can be implemented as a software or a hardware.
We can get the following advantages:
- Commodity classification into Harmonized System (HS), to ease the workload of economic actors and allow better compliance and reliability, both for customs and for the private sectors.
- Image analysis of maritime containers to improve efficiency of cargo inspections.
- Customs control to discover anomalies more quickly and allow controllers to focus on cases of non-compliance.
- Better targeting for the control of risky commercial shipments through real-time data analysis for cargo inspection to detect smuggling or counterfeiting.
- Revenue collection methods, to ensure that duties and taxes are properly calculated and collected at the border.
Artificial Intelligence to help for tariff classification
Customs basis are tariff classification, origin and customs value. Tariff classification is assigning a code to a goods. This customs code has a common international root of 6 digits called the HS code.
The problem is to automatically find the tariff code for goods. Deep learning of customs tariffs, and explanatory notes plus analyzing a huge amount of data for international supply chain can provide faster and more reliable tariff code than human being.
The challenge can be met with a search engine including translates NLP language (product description or keywords…) into customs language to find the tariff code by tackling customs regulations complexity. By using natural language processing and deep learning technology layer, we can get the best tariff code from the product information.
The beliefs about AI
Some experts predict that AI will destroy jobs. There is no scientific proof for that. Indeed, some jobs will disappear and others will appear.
Workers will need new skills in a changing environment. New skills will be important for the labor market. Some jobs will no longer be relevant because technology will be able to perform repetitive tasks faster than human.
Getting AI technology accepted: not easy task!
The challenge for Customs lies in the enthusiasm for adopting this new emerging technology. The changes for AI to take hold in Customs as elsewhere will be met with inertia in current practices.
However, we should be optimistic because innovations always had to face debates necessary for their acceptance within a society. Historically, it’s not the technology that creates the breakthrough but the technological trajectory of a product or service towards the traditional market that causes the breakthrough. Machine Learning or Deep Learning technologies will deeply transform the positioning of customs into international trade.
By Maxime BOADJI, CEO, Yellow Relay & rDuty
April 6th, 2021
