To date, much of this information has been underutilized. To be usable
for customs investigations and risk assessment, the data needs to be
collected and fused together in a single platform, with advanced analytics
and machine learning models needed to parse through it and extract
insights. With dozens of siloed data sources, red flags are easy to miss.
When viewed holistically, those same data points can indicate a pattern of
suspicious activity. Unfortunately, lacking the right systems, few custom
officials can see the holistic view of an individual package, passenger,
broker, vessel, or other entity of interest. As a result, they may miss the red
flags that a shipment should be inspected or that an import/export broker
or shipper should be investigated.
To improve their operations, 44% of customs authorities report they are
currently using big data analytics, AI, or machine learning, and another 33%
intend to start in the near future
4
.
While many customs organizations are already investing in AI and analytical
tools, as is seen in the survey, these solutions often are far too limited
and fall short. Many existing solutions struggle to handle the volume and
velocity of data available, are unable to process both structured and
unstructured data or to utilize data sources that are siloed in different
source systems. Furthermore, there often remains a ‘decision gap’ -
a gap between the insights derived from the data and the resulting
actions that should be taken. Analytics tools and systems often focus
on building models and making projections, but the insights they generate
aren't actionable.
Stage of adoption of big data analytics
No plans
Plans to use
big data
analytics
Using big data
analytics
33%
23%
44%
Source: 2021 WCO Annual Consolidated Survey
7
Data
Challenges
3