Siete 24 Banking Case Study Colombia

Banking Case Study

Siete 24 Banking Case Study Colombia

Overview

The AI surveillance solution was developed to help prevent robberies in ATMs in service by Siete24. The challenge included not only AI-based 24/7 Scylla video content analysis agent but also Face Recognition based access control system that will eliminate identity-theft related robberies as well as cases when the access to ATM vault was forcefully taken. While the first part of the solution would alert the authorities the very second the violator is in the sight of an embedded camera with forbidden objects and/or performing suspicious acts, the second part would serve to eliminate the felony involving personnel that have access to ATM vaults.

CHALLENGE

Detect robbery of ATMs on the early, preventive stage.

Detect specific suspicious objects typically used during robberies.

Detect threatening, suspicious behavior using embedded low profile camera.

Develop the solution to work on small factor embedded hardware to analyze video content on site and send out the alerts only.

Provide face-recognition based Access Control solution for bank employees and service staff that will work independently from ATM main functionality.

Integrate QR-based trigger for initialisation and a stand-alone messenger app to receive clearance, alert failures and enroll a new personnel member.

OUR SOLUTION

Scylla’s modules monitor the CCTV cameras positioned frontally embedded and on the top of the ATM to identify breaches of security protocol and send real time alerts to relevant security personnel.

Scylla Object Detection module used to identify drills, metal rods and other suspicious objects in possession of individuals who are in proximity of the ATM.

Scylla Anomaly Detection and Behavior Recognition used to identify suspicious (anomalous) behaviour of individuals who are in proximity of the ATM and/or wearing a mask/helmet as a disguise.

Scylla modules optimized to work more efficiently on small form factor PCs by various state-of-the-art optimization techniques.

Scylla Face Recognition used to identify bank personnel from a whitelisted database of all employees and grant access to inner vaults or otherwise warn relevant security personnel.