Large volume, complex Regulatory Compliance made simple and streamlined

The Background

Post Brexit, new regulations were put in place for moving goods between the UK, Northern Ireland, and the Republic of Ireland. In response to this, a World Leading Logistics company was required by its large retail customer to be able to report on every consignment going through Northern Ireland from the UK.

The normal procedure would be to submit their customs declarations in paper form before being granted permission to import.

The Trader Support Service (TSS) is there to support this process by enabling organisations to submit their declarations through an online portal, which supports various types of declaration such as ENS Safety and Security Declarations.

There is potential to drastically improve the time taken for data to be processed this would allow more documents to be worked through at a much faster rate. Reducing delays in goods being imported into Northern Ireland whilst also enabling historic data of goods shipped post Brexit to be processed into a standardised format and consolidated.

Goal

The goal is to develop a bespoke mechanism of processing supplementary declaration data from multiple sources that can handle a mass volume of shipment information, and then funnel this through a data pipeline into the TSS platform, as the supplementary declaration portal used by the UK government and freight industry.

Ultimately, the aim is to automate the otherwise impossible human effort needed to create a means of allowing a large retail end customer to view its shipment information in the TSS platform, as required.

The Challenge

Historical data files that contain Customs information of both Import and Export goods will need to be pre-processed in order to migrate the data into a supplementary declaration query system. By doing this the data can then be mapped and enriched with supplementary declaration data from various sources.

Handling these files is typically a time intensive task. Not only will there be a large volume of 3000-4000 historic files to rationalise, there's also an average of 4,000 products/commodity codes per file. With each product line containing then 75 data points, this amasses to a vast volume of data that is required to be processed and written as a standardised format and data compliant single file that is compatible with TSS platform.

The Impact

Implementing these RPA solutions enabled the logistics company to focus on driving their organisation forward from a proactivity standpoint, granted by enabling the solutions to deal with the historic data in a fast and in-line fashion. 

Furthermore, it had a tremendous Return in Investment impact, saving large amounts of time and human effort. These tasks would have otherwise been delegated to operations teams, meaning that these teams were instead able to focus on working with customers and working on processes as opposed to working in processes.

Since the RPA solutions enable the processes to be completed at a much faster pace without the risk of human errors, something that previously was a concern before, the speed and efficiency of delivery have been increased.

The Alphalake Solution

Alphalake AI chose to work with a low-code technology approach, which involved using a Robotic Process Automation (RPA) platform to provide high levels of flexibility when modifying the technical rules across large data files. RPA was used to build a Bot that performs a number of programmed data lookups and edits.

The Automation Anywhere bots are designed to handle any graphical user interface (GUI) based automation between XML files and spreadsheets to create a data processing pipeline which if required, can mean no human intervention is needed.

The bot was able to be easily adapted in accordance with the business logic. This low-code methodology resulted in less time developing the data engineering process when compared to a native code approach. This allowed our DataOps team to work closely with our IAOps team to use the RPA lookup bot in conjunction with our Data Lead, Ravi Theja, who worked across the project, ensuring the continuous understanding and translation of the business need into how the complex data pipeline is managed and ultimately presented to the end customer portal.

As the Automation Anywhere Task bots are very capable of handling various file formats such as XML, XLSX, and CSV, this enabled seamless GUI interaction in order to fetch, compare and integrate data in the desired output format based upon the decision logic.

With regular expressions and parsing capabilities, the task bots provided an elegant solution in dealing with data munging challenges that replicated human intelligence and activities.

Data from different sources and formats such as PDF, CSV, Excel, and XML files needed to be consolidated into one master file, enabling all data to be processed to ultimately match the required format by the TSS portal. The Alphalake DataOps team worked retrospectively through the data from each month of 2021, using the API developed by Alphalake to transfer the information from this single data file format into the TSS platform.

Measurable Value

After automating this lengthy manual process, the bots were able to process five rows/lines at a rate 90% faster than that of a human.

The bots processed a combined 191,454 rows X 75 columns through March, April and May’s data in 638 hours. It has been taking human workers approximately 15,955 hours to process the same amount of data. This equated to a saving of 638 workdays.