Page 58 - Cover Page Sequence-2020.cdr
P. 58
2. Smart deployment: Integrating AI into the jobs flagged immediately to a border personnel, but by
of officers could provide much-needed support in tracking and analyzing data in near-real time, an
the early detection of dangerous people, goods, and AI-enabled system would flag this individual for
activity. Security staffing has a significant impact further investigation by the border security guard
on the resources required to physically patrol to be at risk for human trafficking. AI-powered
border crossing areas. By employing the use of machine vision could also support this in-depth
smart sensors coupled with integrated datasets on investigation by using infrared cameras and blood
migration patterns, crossing activity, demographics, flow pattern recognition systems to recognize
environmental data and more, AI could work deception in an individual. Studies suggest humans
with security organisations to reallocate or deploy are able to spot a lie about 54 percent of the time,
personnel directly to areas of the border where whereas AI has an accuracy rate of over 80 percent.
they would be most effective in that particular Through machine learning and the integration of
time and space. Often it has been seen that the more and more data, these methods are continuing
transborder crimes have a pattern which are linked to be refined.
to several other geo-physical-cultural factors. A
nuanced AI model can provide interlinkages to 4. Logistics and mission planning with purpose:
these factors (external and internal) and help Combine environmental, asset, and historical
us to strategize better deployment, for example operational data to better predict mission scenarios
human trafficking may peak in some months (may to ensure resources are allocated advantageously,
be during months when there is a cyclic recession coordinating between multiple units and
in employment opportunities across the border) operations or task forces. Identify optimal areas
or narcotics movement may peak in some other for camps, and evacuation supply routes for both
months and routes depending on harvest seasons military and humanitarian efforts.
of neighbouring countries or illegal migration 5. Optimum use of existing tech: AI can also be
may increase as result of natural calamity or as a used to enhance existing systems, such as x-ray
reaction to some new legislations (domestic as well scanners for luggage/ package/parcel review. Using
international) – we can be better prepared and computer vision and machine learning algorithms,
effective if the AI models can predict the higher dangerous goods can be more accurately and
probabilities of such trans-border crimes.
consistently identified, reducing the amount of
3. Ops insights: There is so much data being tracked manual labour and possible human error involved
every day, but often its value cannot be unlocked in having to personally examine the scans.
simply because it doesn’t sit in the same place. 6. Suspect identification with accuracy: Use AI to
Integrating previously siloed datasets can reveal recognize patterns of behaviour in an individual’s
connections that a human would not be able data that could indicate suspicious activity and
to piece together–but AI could. AI-supported flag for officer review. Use biometric technology
systems could collect traveller data associated with to accurately match travellers to their travel
a particular person– identified biometrically to documents to improve flow of traffic at security
ensure their identity–who has been entering and checkpoints, flagging only certain individuals who
exiting the country frequently, each time with a present irregularities.
child that has a different name. This might not be
7. Surveillance with integrity: Combine migration
pattern, demographics, environmental data, and
border activity data to provide increased situational
awareness along border crossings and effectively
prepare and deploy security personnel. Use
smart sensors and image recognition to improve
surveillance activities and object detection across
large areas of border crossings to refocus security
personnel’s efforts on responding to unauthorized
crossings and other security breaches instead of
conducting ad hoc patrols.
8. Predictive asset maintenance with confidence:
Use sensor technology and computer vision to
57 Anniversary Issue - 2020 53
th
