Case Studies Blog Careers Technologies
ATM Real-time Remote Monitoring and Alert System Solution
Python, Postgre SQL, Vue.js, Docker, Jenkins, React Native
About Client

CMS is the 5’th largest ATM Cash Management company Globally. They enable commerce – connecting business, banks and people with money. They automate ATM and currency management in India. Their networks and support services ensure money is readily available across all states. They provide Cashiering services for top retail chains to picking up cash from more than 40000 merchants and banking it, they provide a range of services across each stage of the cash cycle in India from currency chests to ATMs to vaults to stores to wallets, to installing and managing Intelligent ATMs, Cash Deposit Machines and Recyclers; they are pioneer in helping change banking in India.

Problem Statement

ATM cloud-based video surveillance using AI is a cutting-edge technology that offers numerous benefits for improving security at ATM locations. By using AI algorithms to analyze video feeds from cameras, the system can detect potential security threats and anomalies in real-time, providing a highly effective means of deterring criminal activity. To date, CMS used third-party vendor services for ATM monitoring system which has limitations and zero control over customizations for additional insights. 

As part of CMS's objective, it aims to build a bespoke ATM remote monitoring system that incorporates multiple features and analytics on video streaming of ATMs PAN India. CMS aims to build a bespoke platform providing below capabilities

  • 24/7 audio and video recording and AI-based monitoring
  • Ticket allotment to user agents based on trigger events
  • IoT Platform enabling no code to low code click-to-build workflows
  • Monitoring of indicators by built-in sensors
  • Automatic notification system in case of non-compliance
  • Multi-tenant platform 
  • Cloud agnostic
Oneture's Role

Oneture being a trusted digital transformation partner of CMS, got involved right from inception of idea in building strategy with leadership team. To target GTM and to prove business value and acceptance of the concept, it was concluded to build MVP of the ATM Remote Monitoring System (RMS).


Oneture worked along with CMS technical team to provide below key features for the scope of MVP:

  • EdgeBox API Integration 
  • Video Analytics based AI Bot
  • Ticket Management
  • Device Management 
  • Workflow Management

EdgeBox API Integration 

An Edgebox is an IOT device that combines all underlying components like Cameras, Sensors, Electrical Units (AC, Lights, Smart Switches) into a single unit and is responsible for controlling the exposure of the components over internet. 

Edgebox Connecter – An Edgebox connector exposes number of OEM functions / methods for end user to integrate with any external Application Programming Interface. 

Edgebox Connector API – An Edgebox connector API interacts with the functions / methods offered and exposed by OEM.

Edgebox Integration Criteria:

    • The Edge Device API integration should be independent of the device make
    • The Edgebox API will be accessed from RMS API using a wrapper.
    • RMS APIs must be load tested as  N number of Edgebox APIs will be consumed at the same time across the ATM centers

Device Management 

Device Management will allow user to Add/ Edit/ Update/ Remove devices.

System to offer a provision grouping Devices connected to a single site for better identification.

System to offer categorization based on the respective device category

Workflow Management Console for Ticket Creation

A workflow is typically a process, designed as a combination of interrelated User Activities / Automated Activities to achieve a certain objective and which can be triggered manually/automatically by any external or internal event.

The workflow represents an integrated workflow management console provided within the RMS MVP solution. It defines SOP / Business Use Case flow and depicts the same as a graphical representation of activities. From a design point of view, it should be a graphical user interface (GUI) that will allow a user to define (drag, drop, configure & publish actions) the process flow for any alert configuration. 

The workflow will have the following capabilities:

    • Visual / Graphical flow of the process.
    • SLA Based
    • Configurable 
    • Process Change Management
    • Activity Indicators 
    • Ticket Management Integrated 

Typical Workflows identified for providing out of the box support in RMS system are  ATM glass break, heat raise, temperature high/low, panic, fire, atm floor, atm chest, panel door, and vibration.

An example of ATM Glass Breakage Workflow


Ticket Management System

As described in the Workflow Management Console, Ticket Management System will follow whatever is defined under the workflow. 

Any Automated / SLA Triggered / AI Driven event can invoke a Ticket in the System. The ticket will follow its lifecycle as per the respective definition under its workflow as part of SOP. 

AI BOT will always attempt the workflow to force close any given activity if the scenario fulfills the AI match. When the case doesn’t fit into auto closure it will be assigned to the respective Agent to analyze the root cause of the problem and close the case with an explanation.

AI based Bot

The workflow will be driven and continuously monitored by AI BOT (Built for injecting AI).

Soon as the workflow will be triggered, AI BOT will simultaneously run check if it the process can be completed without Agent intervention.

If all process-based checks are met, it will mark the activity as completed else for the case outside of the pre-defined learning, it will assign Agents to resolve the Ticket. 

Realtime Video Analytics based Bot

The solution will support Native video analytics running on edge device to trigger tickets and necessary alerts based on pre-defined events as following:

    • Helmet / face cover
    • Loitering / People overstay
    • People count / crowd
    • Blackout / camera vision

Oneture will also be exploring AWS IoT and video analytics services leveraging AWS IoT Greengrass and AWS Kinesis with AWS Rekognition for real-time video analytics. 


  • Frontend: VueJs (Web), React Native (Mobile)
  • Backend: Python Django, RabbitMQ, PostgreSQL
  • IoT: MQTT, SIA DC09, RTSP camera feeds
  • CI/CD: Docker, Jenkins, TFS
Value Delivered
  • Live ATM monitoring PAN India
  • Accelerate Dispute Resolution – One common interface for all events and devices
  • Integration supporting Edge device of popular protocol