South Central Ambulance Service: Predicting Ambulance Failures with Data - Ctrl Alt Lead podcast

SCAS is combining data sources to understand failures, improve maintenance and keep clinicians on the road

For an ambulance service, availability is more than an efficiency metric. Failures leave first responders unable to reach patients; here's how one organisation is using data to minimise them.

South Central Ambulance Service (SCAS) NHS Foundation Trust transports patients across Berkshire, Buckinghamshire, Hampshire and Oxfordshire: an area of vastly different driving conditions ranging from inner cities to the New Forest, putting high stress on vehicles.

In the latest episode of Ctrl Alt Lead, SCAS' head of data, analytics and insight Mark Green and senior management information analyst Vivienne Parsons explain how the trust is using vehicle telematics, machine learning and analytics, alongside geospatial data from Ordnance Survey, to better understand and prevent failures that can take expensive vehicles off the road - and, more importantly, delay clinicians responding to patients.

The scale of the challenge is imposing. Telemetry arrives from vehicles every 10 to 30 seconds, generating millions of rows of data each month. But by combining this information with details like road gradient, camber and surface, the team can investigate not simply when something went wrong, but the conditions surrounding it.

The approach has produced unexpected insights. Data analysis has shown that the varied road conditions in SCAS' area put different stresses on vehicles, opening up the possibility that rotating vehicles between locations could extend their working lives.

“Southampton versus the New Forest, they’re very different road networks, and equally the A34 has got some very high gradients on it. So actually, do we move the vehicles round because one isn’t suffering in the same way as the other?” said Vivienne.

“Does that make it last longer? That’s something we’d never considered until we went through our work."

With each ambulance costing as much as £200,000 to put on the road, any extension is worthwhile for the NHS- and charity-funded organisation.

The project has also highlighted the importance of off-time, with insufficient charging between shifts contributing to equipment failures.

More effective maintenance is only one potential application, though. Mark thinks AI could eventually help SCAS extract useful information from the huge quantities of data it generates, including the assessment process used to determine how to respond to patients.

“What’s really key is getting to the details, understanding what response a patient needs as quickly as possible and then responding to that appropriately,” he said. “With the amount of data that is created... using AI to deliver that insight is absolutely where we should be going.”

Vivienne believes the opportunity is as much about time to act as it is faster analysis. At present, she said, sifting through SCAS's datasets can take weeks. AI could instead provide teams with “more time to look at the why and come up with a solution and a plan, not just crunching the data."

To access the full podcast watch the video, use the web player, or find Ctrl Alt Lead now on Spotify and Apple Podcasts.