While datacenters (on-premise and in the cloud) are now a standard approach for IT operation management and IT service management to unlock value and support businesses, companies are increasingly interested to apply this concepts to their internet of things world. Based on the type of sectors, use cases could be different:
- For a chemical industry: improve the maintenance process while having visibility on all factory devices
- For an Energy industry: enable and improve quality of Field Service Management for wind turbines while using digital twin concept
- For a car industry: improve customer satisfaction while responding proactively to maintenance requests
- For a Bank industry: Standardize provider relationships and processes while keeping an eye on ATMs health
- ...
All of these are good examples to expand the Service Management capabilities to areas that may not be covered yet, but how to automate and industrialize the Operation Management side?
In this article, I would like to share the bank ATMs use case and describe how we managed those devices on an Operational perspective in a POC approach.
What are the challenges?:
First of all, Bank ATMs are usually in a huge number. We can have several thousands of ATMs across the world for a particular Bank organisation. Having the visibility of all of them and applying service management processes is difficult when or even before an issue occurs.
When issues occur on ATMs, it can take a long time identifying them and restoring the ATMs which leads to negative impacts on brand and branch reputation. Assessing the performance of ATMs and act proactively is a challenge.
Most of the time, knowledge on ATMs is owned by its specific provider or ATMs experts. Bank usually addresses issues to the provider/experts and wait for it to solve issues or to improve service quality. Without enough insight on ATMs behavior and performance, Banks can't speed up the communication and provide feedback to the provider/expert in order to improve the service quality
What is an ATM?:
ATMs devices are complex and can vary from one Bank organisation to another.
Here what we need to understand from the industry and what the most interesting part for our use case is:
- The vast majority of ATM use a Microsoft Windows Operating System (in our use case, they are windows 7 devices). Our discovery capabilities are then adaptable to this context
- Devices that are usually exposed to customers or the ones that manages ATM tasks can be connected through simple USB connection. This devices can have problems and undergo some outages (example: card reader is not working and customer doesn't know it yet).
- XFS is the common protocol used by ATM to interact with them. But in order to have better flexibility, some organisations have evolved this protocol to use standard ones like REST WebServices or windows based communications (PowerShell, WinRM, WMI...)
What we achieved:
IT Operation Management (ITOM) gathers several practices into ServiceNow platform among which:
- The Discovery solution that discovers in an automated and industrialized manner the devices present in a specific network area
- the Event Management solution that gathers monitoring events from any type of source and that applies machine learning on them to assess devices and services performance.
- An Orchestration solution with Flow Designer & Integration Hub modules to enable automated processes in a simple way, for example an auto-remediation use case on a broken service.
All these practices have been adapted to the ATM use case
Implementation:
1) ATMs have been discovered automatically and all its related USB devices. Business services have been created for each of these ATMs in order to measure its performance. We've decided to implement a new "Windows ATM" class with an "ATM devices" class related to it. We've then attach specific processes to them
A standard architecture has been used for this:
2) The XFS based events from ATMs have been translated into JSON based format and sent to the ServiceNow platform.
All these events were then bound to the specific ATMs or its USB devices which enabled us to visualize their health on the Operator Workspace:
3) Finally, We've created several remediation rules based on the type of event coming into the platform. This enables ATMs owner to trigger some basic processes to execute a first step remediation procedure. For example, the operator could trigger a reboot of an ATM if a specific "102" type of event came into the platform
Results:
The main objective of this implementation is to improve the insight and feedback provided to the ATM provider to ensure ATMs availability and performance. The Service Owner workspace enables ATM owners to compare their performance and identify the ones that went into outages more frequently and the trends of event leading to a particular outage. The different branches could be notified on the potential outages proactively or in real time.
Having a CMDB up to date with the latest information of the ATMs regarding upgrades, installations, traceability on change, incident and problems also enables faster decision making and remediation.
