The whys and hows of proactive ATM maintenance
by Jim Tomaney, consultant, Paragon Application Systems
When an ATM fails, it can cost the operator a significant amount of time and money, not to mention the damage done to the operator's reputation. With social networking apps at their fingertips, customers who find an ATM they depend on out of service can tarnish a brand with just a few keystrokes.
To avoid these headaches, financial institutions and third-party providers of ATMs should consider a move from reactive to proactive maintenance.
Reactive maintenance — too often the default — deals with problems only after they occur.
Proactive ATM maintenance, on the other hand, focuses on identifying and solving problems before they materialize.
But in order to ensure that proactive ATM maintenance is worth the investment, ATM providers must test the results to ensure they can accurately and dependably prevent failures.
As previously mentioned, reactive maintenance of ATMs can require a considerable investment of time and money.
If the organization isn't aware of a problem until it occurs, it might be forced to dispatch an engineer to the scene, which is costly and disruptive to the regular schedule.
And, of course, the longer an ATM is down, the more business the provider loses. Consumers have grown increasingly impatient with slow and inconsistent self-service experiences, and become easily frustrated when the ATM they visit is down.
Given these facts, taking a proactive approach to ATM maintenance has some advantages over the reactive model that are worth considering.
One type of preventative maintenance is typically performed by the engineer who has been sent to an ATM in response to problem. While the engineer is at the ATM, he or she will carry out preventative maintenance in an attempt to prevent future issues.
Engineers generally will check the top 10 known fault conditions while they're at an ATM, but their schedule is usually too full to conduct many additional tests that might be useful.
Fault testing can be a largely manual process requiring a good deal of time, however it's certainly better than no preventative maintenance at all.
Another approach to proactive maintenance might be more attractive to organizations that wish to improve efficiency while lowering operational costs.
This approach uses a monitoring agent onboard the ATM that automatically screens the behavior of the device and records errors and warnings. When a fault is detected, the monitoring agent can execute local remedial rules to attempt a fix. If the problem cannot be resolved in this way, the system can forward the events to a central monitoring server for action.
The monitoring server analyzes data received to identify any patterns that might have led to the failure. By working backwards through the data to determine the sequence of events that occurred at the ATM before the failure, the system can potentially recognize the symptoms of a particular type of failure.
If the system is aware that a pattern of X, Y and Z warnings might lead to error A, it can apply a rule set for a potential situation and alert the operator to the need for preventative action — for example, to replace a part that is expected to fail soon.
Using this approach means conducting maintenance on machines when they need it. The automated and adaptive nature of this approach also means that organizations can easily and quickly respond to changes in a machine.
The problem most ATM providers encounter with this approach is that the only way to verify that an event or pattern indicates a particular outcome is to manually test the machine in the lab using data gathered from machines in the field.
This is not only burdensome, but also time consuming, as it might take two to three months to thoroughly analyze the data. Only then can the operator accurately determine which errors can be avoided, how often a rule works, and how often the error still occurs despite intervention.
From there, the tester can make a recommendation to test the rule again, or to test other conditions instead.
To eliminate the weaknesses in a traditional, lab-based approach, ATM monitoring rules should be tested in a virtualized hardware environment, where countless complex failure scenarios can be creating using scripts.
A script can run different events, recreate complicated situations and test how the monitoring solution detects these events, executes the rules and maintains the ATM.
Proactive ATM maintenance and monitoring can offer a clear competitive advantage to an organization that manages ATM networks and wishes to avoid the cost, expense and reputational damage associated with unplanned outages.
Using hardware virtualization and test automation, an organization can develop and implement sophisticated monitoring rules to ensure that ATM availability remains high and ATM users remain satisfied.
|Jim Tomaney is a consultant for Paragon Application Systems, a provider of software testing solutions designed to determine the reliability and integrity of electronic payment systems. More than 600 institutions in more than 90 countries use Paragon products.|
Companies: Paragon Application Systems