zatletic - Fotolia
Operational and logistical issues tend to dominate the operation of service delivery organisations such as Australia’s Royal Flying Doctor Service (RFDS), whose employees work round the clock to deliver medical emergency services to residents across 71.5 million km2 of the country.
It is not surprising, then, that it was for a decidedly administrative purpose – analysing operational logs – that the organisation implemented the Splunk Enterprise big data analytics system.
The idea was to scour logs for network alerts and security-related anomalies that might highlight issues that needed addressing. But it didn’t take long before business teams began to get other ideas for the system, according to Adam Ind, RFDS IT manager with the central operations team, which covers the regions of South Australia and the Northern Territory.
“We quickly grasped how easily we could create a query and get answers and insights into the operation,” says Ind. “Our reseller [Adelaide-based SecureWare] suggested it would be quite powerful for looking at our business operations as well.”
Early experiments proved that the analytics environment was not only powerful, but rapidly so. After loading masses of historical flight-log information from its Sabre AirCentre Flight Explorer tool – which logs the movement of the 63 RFDS aircraft (15 of which service Ind’s central operations jurisdiction) between its 21 bases – the flight management team began to spot patterns in its data that they had never before noticed, or even thought to look for.
“It was quite amazing to start with, just having a bit of a play around to see what insights we could get and what patterns we could see,” says Ind. “Just having easy access to that data gave us many opportunities that would otherwise have been a very big, serious undertaking – and we were able to feed that back as a proof of concept in the organisation.”
That’s when the marketing department got on board, in a serendipitous turn of events that helped cement the system’s future in the RFDS.
Marketing, it turned out, had been looking for an innovative fundraising campaign that would help the organisation engage the public in a new way. With the Splunk system providing easy analysis of the movements of the service’s planes, the idea was hatched for what eventually became the RFDS’s biggest marketing campaign ever.
Launched in late 2013, the Buy the Sky campaign allows the public to “buy” a patch of airspace along the service’s flight paths, virtually travel through those areas, and get regular updates when planes fly through those patches.
Those updates are provided in real time, thanks to the analytics engine that is constantly running against the ever-growing Flight Explorer dataset. It was a revolutionary service for RFDS – but it was only the beginning of the value that analytics was set to deliver for the 87-year-old organisation.
Analytics take flight
Over the course of the following months, Ind and his team began looking for other datasets to feed into the analytics platform. Avionics data from the aircraft, for example, is collected on a continuous basis and stored, but had never been utilised in an analytics context.
Charts of planes’ movements, maintenance reports and data from a range of other specialised systems provided by aircraft manufacturers were thrown into the mix, and a much more detailed picture of the RFDS operations began to emerge.
“We weren’t really doing that much with the data in the past, but when we ingested the log files and integrated the different databases, we had queries at our fingertips and could manipulate the data and output it in that format,” says Ind.
“We were able to do things like create time charts showing the number of aircraft in flight during any particular hour over the past week, which gives us an idea of how activities ramp up and ramp down overnight.”
That sort of information has significant value for RFDS administration, which, like any airline, must constantly watch costs as it seeks to optimise routes flown by its aircraft and minimise ongoing costs in areas like maintenance and staffing.
Read more about big data in Australia
- Data from analyst firm IDC shows that big data and advanced analytics in Australia could be about to accelerate.
- Big data is like digging for gold, except you must keep every single rock. Computer Weekly unearths some of the main challenges facing those mining data.
- Australian businesses need to change their attitudes towards data scientists if they are to unshackle the benefits of data.
Different combinations of data are being evaluated for their value in driving an activity-based costing regime. And relevant reports and dashboards are published on the RFDS intranet and projected onto digital signage throughout the organisation’s operations area, providing all staff with unprecedented visibility into its everyday operations.
Building data value in Splunk
Regular dumping of onboard systems data has also turned up new uses for the analytics that were never envisioned back when the Splunk licence was first implemented.
For example, the analytics environment is now being fed information from Wi-Fi-enabled temperature loggers in the drug fridges and drug bags used to keep temperature-sensitive medications from spoiling on the long trips to places where the mercury regularly passes 40ºC (104ºF). Logs from that hardware are automatically fed into Splunk and alerts transmitted when an exception is detected – providing a new degree of governance that had been difficult to achieve in the past.
“For anything that’s outputting any regular data, my immediate thought now is always ‘what can we do with this data in Splunk?’,” says Ind. “Most of our specialised trending systems only let you analyse snippets of data at a time, but because we are actively indexing all of our data we can look at those trends over time.”
Many of the datasets in use are actually quite small, but when put together they paint a detailed picture of change over time. “I’m almost surprised at the value we’re able to get given the size of the data we have,” he says. “When I consider that we are tracking every aircraft in our fleet, plus the data coming from their onboard avionics systems, I’m surprised at how little storage is actually required.”
Recognition across the business
Armed with reports from the analytics system and the promise of even better information, Ind has secured newfound credibility with the business arm of RFDS, encouraging them to join the conversation about what data should be integrated into the system into the future.
The central operations team has also attracted attention from other RFDS operational areas that face similar challenges as they come to appreciate the value that the analytics has provided the staff in Ind’s region.
“It’s facilitating a conversation within the organisation to try to establish what the questions are that people don’t have an answer for,” says Ind. “Pretty much everyone in central operations has exposure to the data in some form.
“And while it’s hard to quantify the value we’re getting, its value is fantastic in terms of the improvements that we were able to make in terms of quality or responsiveness. It really is invaluable to us.”