CORE: Customs has high expectations of data dashboard CRIS

A smart tool could enable Customs to effectively free-ride on the information systems of companies. This would bring benefits to these businesses as well.

​In principle, data pipelines of companies are a valuable source of information for Customs administrations. The proviso for this being that the service can easily plug in to such a digital flow. And that the input thus acquired for risk analysis and risk selection can quickly provide insight. What may prove to be a useful tool in this, is the Customs Real-time Information System – CRIS, for short. This smart dashboard, which is still being modified, can soon also generate profit for trade and logistics.

Customs’ response to what other governments, companies, educational institutions and research centres have been working on within CORE, thus Mitchell Out’s definition of CRIS. “Within the framework of ongoing sub projects, a dashboard for data exchange had already been developed by an external software supplier. We have taken useful functionalities from this and poured our own sauce over it”, says the information management specialist at Dutch Customs’ National Office, who, as project manager, is responsible for the development of the new application. “Our tests namely showed that although the original version met the technical requirements, it did not really improve the quality of our work. Our employees were given yet another screen, which meant additional manual intervention. While, in fact, we were seeking an integrated, less labour-intensive solution. We seem to have found this in CRIS.”

360-degree overview
During CORE, Dutch Customs focused on the Entry process, more specifically on what is known in jargon as the ‘orange channel’: declarations that have been ‘ejected’ by the system for touching upon a risk profile. Something may not be quite right about the associated consignments, and they may be eligible for further inspection. It is up to the risk assessors of the service to determine whether this inspection will actually take place. Out: “They traditionally base this mainly on information provided in the declaration. If needed, they can consult all sorts of other data files – think for example of the Commercial Register of the Chamber of Commerce. In so doing, they each time consult individual databases, something which is relatively time-consuming. Moreover, which databases he or she will consult depends heavily on the experience and the mindset of the individual employee – there is no standardised working method for this. Thanks to CRIS, we can in due time offer our people automated additional sources in one single window – very convenient. This creates a 360-degrees view of the most important information that is available, which allows for an easier, better and more uniform manner of assessing the ejected declarations. We try to organise it in such a way as to be able to easily find out that goods are probably less risky than the system initially suspected.”

Fewer undue delays
The data pipelines of companies will also have to feed the customs dashboard. With real-time information from the business itself, on trade documents and tracking & tracing of goods for example. Out explains that these data will also greatly contribute to de-risking. “Suppose a declaration refers to a delivery from one forwarder to another, and it is not known which sender and receiver are behind it. This can be sufficient reason for our system to eject that declaration. If a risk assessor can then check the pipeline of the specific logistical chain to see that these are reliable market players known to Customs, the conclusion can then be to waive the need for the goods to be inspected. On the one hand, this will thus decrease the number of communications – emails, telephone calls – between our service and companies, as there is less need to approach them for missing data. After all, we can now retrieve this information ourselves if needed. On the other hand, the number of false positives*shall decrease, which leads to less unnecessary physical goods inspections and therefore to less unnecessary logistical delays. This thus allows for more time and capacity for parties that conduct themselves in a non-compliant manner and that deserve our full attention. By doing so, we optimise our supervision.”

Enforcement vision takes shape
“Our enforcement vision ‘Pushing Boundaries’ distinguishes three layers in trade, all referred to with a different colour”, Out proceeds. “The difference lies in the extent of familiarity and reliability of the companies and chains that act within them – from our perspective. At the top, is the so-called yellow goods flow, which consists of logistical links that are considered compliant as coherent chains. Such a supply chain requires significantly less physical controls on goods. It usually suffices to check whether the parties involved manage their logistical and business processes and comply with procedures –  something which is a lot less laborious for us.** Within CORE, we tried to test a number of these potential yellow flows in a test environment. The entire Kenya route of Royal FloraHolland – from grower to florist – has been mapped out for instance. The benefit for us is that we now know when and at what places within such a process important information and documentation is created, and thus where our points of measurement are to be found in the associated data pipeline. CRIS enables us to plug in at the right time, and track to see whether everything proceeds as it should. All and all, these types of pilots contribute to the formation of yellow flows, which we can both monitor as well as facilitate increasingly better. Our vision of the future customs supervision gradually gains shape.”

Yet Out hastens to add some reservations to his promising picture. “Do not rely on a high pace for such developments. CRIS was a proof of concept for us, which means, among other things, that we used old information sets during the test path. Before converting the system to the production environment – and thus with up-to-date information – certain steps still need to be taken. Supervision on the basis of dashboards and data pipelines is new to us: we cannot yet exactly determine the effect on our risk selection. Nor can we estimate just how quickly this working method shall be adopted by the organisation. After all, it takes time getting used to. We shall therefore approach its implementation cautiously and structured – in phases. We will first run the following pilot projects, with employees from the Customs National Tactical Centre, divided between Rotterdam and Schiphol. These are expected to start at the beginning of next year.”

The transition from test mode to supervisory practice has its challenges, Out continues. “It is important, for example, to determine in advance which data we actually want to retrieve via the pipelines. This in order to avoid being overloaded with information that is of little or no use to us. As an example, within the framework of the ongoing CORE project, Maersk/IBM provided us a lot of detailed information about logistical movements, such as actions on the quay, which are of limited importance to us. We must gain more experience and engage with business and industry to keep things manageable.”

In spite of the issues and the points of concern mentioned above, Dutch Customs will, as stated, proceed steadily in creating CRIS. “The organisation has given the green-light to continue the chosen course”, thus Out. “This is mainly because of the fact that colleagues who have participated in CORE and have become acquainted with the dashboard are very enthusiastic about it. They refer to it as promising and look forward to get to work on this in their daily work. Furthermore, we also see possibilities outside the Entry process. In principle, we can deploy this instrument for all work flows that are currently still faced with a lot of customer contact about declarations that run throughout our systems. Examples include Import, Transport and Customer Management. The effective free-ride on the information systems of companies equally brings benefits.”

What is more, the use of data pipelines by the market could well take off in the coming years. Out: “We note that more and more parties are setting up such concepts, and asking us whether we could and would like to participate in them. This does not surprise us in the least. A data pipeline is primarily a business case which involves commercial interests. Customs constitutes a potential gain and is therefore a factor to be considered. In the end, entrepreneurs aim for a more predictable chain, also in terms of possible inspections done by our service.”

AI: the next step?
Even though CRIS is still in its infancy, Dutch Customs is already looking into possibilities of fine-tuning the system. Out: “To this end, we consider whether it is worthwhile to automate certain parts of the process of risk selection over time. You could wonder whether in the future a civil servant shall always be needed to interpret and assess the information on the screen. In some simple cases, de-risking could just as easily be done by software. The platform on which the dashboard runs, would in any case support the deployment of artificial intelligence. However, apart from the technical aspects, we shall first also need to look at the procedural and legal aspects thereof.”

* Goods that have been inadvertently selected by Customs to be physically inspected.

** The criteria for the yellow flow have been met when the source data are reliable and the physical integrity is guaranteed during the transportation. There will then still be customs controls, however, these principally take place at the beginning or end of the chain.

Share this post