Transactions & Device Log Analysis for Performance Improvement

[Audio] Transactions & Device Log Analysis for Performance Improvement ATM Pro Team Bokka Srinivas Rao Manoj Sadhu Sreeharibabu Mooda Marka Sai Krupa Subramanyeswari Anusha Kornapati Madhusmita Biswal NCR Atleos Confidential.

[Audio] Problem Statement Transaction Performance Log As different Transactions on different ATMs are taking different times to complete, we are trying to identify the root cause for these issues by analyzing the logs. Note: The required logs are collected and synced to the server on a regular interval. Oh Nooo!!!! Why is one transaction taking so long?? NCR Atleos Confidential.

[Audio] Insight to the Problem Transaction Log Any transaction should follow the allowed time frame. Device Log After series of repeated transactions, if the memory usage reaches the available capacity, the system will hang/crash. Any Deviation from the allowed time frame, will cause longer time to complete the transaction..

[Audio] Proposed Solution Transaction Log To identify the root cause, we need to collect the transaction logs and analyze the transaction which consumed more time to complete. For this, an XML/JSON file is created and placed at XXXXX to record all the transaction logs. This file is uploaded and synched to the cloud server at the regular intervals, where it is analyzed to troubleshoot the transaction performance issue. Device Log To identify the root cause, we need to collect the device logs and analyze the device parameters: memory usage, CPU performance etc. XXXX For this, we will use the above XML/JSON file . The Synched file from the cloud server is analyzed to troubleshoot the device performance issues. Corrective Actions (To be performed in sequence until the transaction/device performance is fixed) 24 Volt Reset ?? Check the applicability on other ATM variant Diagnostic Entry ATM Restart Fix Device specific performance issue State Specific performance issue.

[Audio] Deductions From Log Analysis Parameters We are targeting to calculate the following parameters using the logs: Best transaction time Worst transaction time Set an allowed interval for transaction duration based on best and worst time frame . The allowed interval is considered as base line. Live transaction data (Includes each transaction time taken, memory counters for each 1 minute ) are retrieved to an XML/JSON file. Flags are to be raised in case of deviation from the interval. For example, if the allowed interval is from 1 minute to 1 minute and 30 seconds, and one transaction takes 1 minute and 50 seconds to complete, the alerts must be raised stating deviation from the allowed interval. Host is intimated about the slowest module. NOT COMPLETED, NEED TO ADD MORE DATA.

[Audio] Screenshot for Output. Screenshot for Output.

[Audio] Business Value Operational Efficiency Monitoring ATMs with real-time data logs allows for the identification and resolution of issues such as prolonged transaction time, high memory usage and performance degradation before they impact customer service. Customer Satisfaction Log information from ATMs can help understand customer behavior and preferences, which will help in reconfiguring the existing ATMs variants and deciding on the types of services offered. Cost Management Detailed analysis can help in speeding up the problem resolution. We can also create support tickets which can be resolved before customer raises it. Corrective Actions Predictive Maintenance Advanced analytics capabilities enable predictive analysis of withdrawal trends and ATM behavior, which can inform strategic decisions By analyzing historical data and real-time inputs from logs, predictive maintenance can forecast when an ATM is likely to encounter a fault. This helps banks schedule maintenance proactively, reducing downtime and ensuring better availability for customers.

[Audio] Thank you NCR Atleos Confidential. Thank you.