IoT Based Analytics in ATM Facility
Background:
Have you ever wondered how IoT & Analytics together can add value in Banking World?
In this blog, we are focusing on how IoT & Analytics technologies together can transform to have ATMs into a Smart ATMs.
There are many real-world examples where Internet of Things are being implemented at enterprise level. These examples are ranging from various verticals, including Automobile Industry, Manufacturing, Power & Energy, telecom, banking etc.
This blog focuses on explaining value proposition in Banking Vertical especially in ATM Facility. An ATM in India requires Rs. 35,000 a month on an average on operational and maintenance expenditure, with the average cost per transaction, excluding the security guard’s income, at Rs. 10-25.
Pairing IoT with analytics enables banks within ATM facility help banks analyse sensor data streaming from Door Lock Sensors, Temperature & Humidity sensors, Light Sensors, Smoke Sensors etc. are being captured and analysed to cut costs by saving powers, curbing theft incidents, & managing other incidents like – Card Reader Malfunction, Receipt stuck in the machine, Stuck Cards in the machine etc. These sensors are already available in all ATM centres. Highlighted in the below image:
Data Elements & Sensors in ATM Environment:
Figure 1: Data Elements & Sensors in ATM Environment
As mentioned in the below solution architecture, there are four various modules which are required to be implemented to have IoT based analytics using SAS.
Figure 2: Solution Architecture
- Data Source: Each bank & ATM facility will have to have data available and sensors available in place to fetch the data using ETL operations or Streaming mechanism.
- Data Management: Various SAS Tools like – SAS Data Management Standards for enterprise data management, routine transaction data management & customer history data management & monitoring.SAS Event Stream Processing Server are required to be able to connect the mentioned data sources including Enterprise Data, ATM Transaction data, ATM Sensors data, Social Media data etc. Data will be processed & standardized & stored in Data Warehouse using various database systems. These data, further be, used for Analytical modelling & Reporting needs.
- Analytics in Action: Basis the data availability in DwH, various Analytical models are required to be created to achieve the value propositions for IoT based analytics (mentioned in the appended section of the blog).
- Deployment: Once Analytical models are created, it will be deployed on the production site to train & validate model basis the regular data streaming in the system. Model output is usually presented in the Graphical Dashboard form for the end-users to consume and make valuable decisions on the fly. Various SAS Tools are equipped enough to do the exercise, such as – SAS Visual Analytics is used for MIS Reports, SAS Event Stream Viewers for live sensor streaming data & threshold breach alerts. SAS Visual Text Analysis is used for any customer feedback associated with individual ATM machine.
Value Proposition for IoT based Analytics in ATM facility – Using SAS
Asset failure monitoring & alerts – any incidents like high degree of power usage, smoke sensors to detect the fire occurrence, monitor temperature & humidity level in server room upon crossing a threshold. Reducing service incidents to a minimum and maximizing ATM up-time in a cost-efficient manner.
IoT based analytics Stream Viewing can improve ATM up-time –using POS Terminal’s performance statistics CPU utilization, Memory Usage, High processing hour analysis, which can help understanding the patterns of terminal usage.
Web Based Dashboard – With graphical reports, maps & other KPI monitoring such as – cash stocking & inventory management, Tracking Hourly, Daily, Monthly & Annual Cash Withdrawals.
Voice of Citizen – in terms of grievance posted about the ATM facility, or any social media contents added on various social media platforms & sentiments around the ATM facility whether working properly, not working properly at any given time.
Predictive analysis – Prediction can help suggest cash refill within certain area & targeting specific zones where footfall traffic is high & can help identify idle time in transactions.
Out-of-Order Prediction – IoT & Analytics can help preventing cash-tripping, physical damage to machine, burglary detection & alert generation; predict outage due to cash shortage in the Machine.
After connecting to the server, the data sets can be able to upload in the SAS vi and the uploaded data sets are been seen in the Data Object as shown in the below image.
For more information, please contact us
Written by,
Nupeak IT Solutions LLP