Across rail networks globally, turnouts remain one of the largest contributors to delays, maintenance costs and operational disruption.
At the same time, rail networks are becoming busier, service frequency is increasing, and access windows for maintenance are becoming increasingly constrained.
Rail operators and infrastructure managers are facing several challenges: enhancing turnout reliability, extending asset life, minimising time on track, and reducing whole-of-life maintenance costs.
Salix, in partnership with KONUX, is introducing Artificial Intelligence-driven predictive turnout monitoring to Australian rail networks – enabling operators to shift from reactive maintenance to data-driven asset performance management.
“As networks move toward higher utilisation and tighter performance expectations, predictive maintenance is rapidly becoming a strategic capability rather than an operational improvement,” said Mark Fulford, Chief Executive Officer of Salix.
Predicting failures before they happen
KONUX Switch is designed specifically for turnouts: one of the most failure-prone and operationally critical assets in rail infrastructure.
By combining Internet of Things (IoT) devices, AI, and deep rail expertise, it enables early failure detection, precise diagnostics, and data-driven planning. The result: fewer disruptions, lower operating expenses, and longer asset life.
At its core, KONUX Switch continuously analyses the health of critical turnout components such as the switch, crossing, point machine and trackbed – providing early warnings and actionable insights.
Compact, self-contained sensors are installed in under 15 minutes and operate with battery life exceeding five years. Once installed, the system continuously analyses asset health and provides early warning alerts, enabling maintenance teams to intervene before failures occur.
The AI-driven platform has been trained using more than 500 million train traces across more than 10 countries, supported by multiple predictive models. The result is a highly refined predictive capability that delivers up to 90-day forecasting of potential turnout failures.
Operators gain access to a cloud-based dashboard that delivers asset health rankings, early warning alerts, predictive failure forecasting, maintenance validation insights, and network-wide visibility. This comprehensive view enables maintenance teams to plan interventions at the optimal time with the right resources, reducing reliance on reactive maintenance while minimising operational disruption.
Predictive turnout monitoring
Rail operators adopting predictive turnout monitoring are typically targeting three key outcomes:
Improved reliability
Early detection of turnout degradation reduces failures and improves network performance. As turnouts are one of the most common causes of delay minutes across rail networks, this makes reliability improvements particularly valuable.
Less time on track
Predictive maintenance reduces emergency callouts and unnecessary maintenance interventions. With access windows becoming more constrained, reducing time on track is becoming a major operational priority.
Longer asset life
Data-driven maintenance enables operators to intervene before degradation accelerates, extending component life and reducing lifecycle costs.
Together, these outcomes deliver measurable improvements in network performance and operational efficiency.
Proven performance
KONUX Switch has been deployed across more than 7000 units globally, with adoption by major rail operators including Network Rail in the United Kingdom, Deutsche Bahn in Germany, Transport Scotland and Infrabel in Belgium.
Fulford said these deployments have demonstrated operational benefits, achieving up to a 50 per cent reduction in delay minutes, a 40 per cent decrease in repair downtime, a 30 per cent reduction in operating expenditure, and up to a 20 per cent extension in asset life.
On Network Rail’s West Coast South route, where more than 500 devices are in operation, the business case has been driven primarily by delay prevention. The deployment has delivered a £11.7 million net benefit over five years, achieved a payback period of just one to two years, and realised a benefit–cost ratio exceeding 4:1.
“These benefits began accruing immediately following deployment, demonstrating the value of predictive monitoring as both an operational and strategic investment,” said Fulford.
Detecting failures before they become disruptions
In a deployment with Transport Scotland, 131 monitoring devices demonstrated a 39 per cent improvement in performance following installation.
At Winchburgh Junction, an alert identified a developing rail crossing defect that would likely have resulted in a failure. Early intervention allowed maintenance teams to undertake planned renewal works, preventing disruption and improving asset life, a joint study between Transport Scotland, Scotland Railway and Network Rail UK showed.
In a comparable junction that was not monitored, a similar defect caused a total of £1.35m in delay and ratification works.
“In environments where even minor failures can cascade into major network disruption, predictive monitoring is enabling operators to move from reactive maintenance to proactive asset management,” said Matt Weingarth, Director at KONUX.
Reducing unnecessary maintenance
Predictive monitoring is improving maintenance efficiency and reducing unnecessary interventions, as demonstrated by a trial with Infrabel in Belgium, which analysed 33 tamping interventions across 12 turnouts.
It found that 40 per cent were unnecessary, only 15 per cent were fully effective, and some even accelerated degradation.
“This highlights how validated maintenance insights can reduce unnecessary work, optimise scheduling, improve intervention effectiveness, and minimise track access requirements,” said Weingarth.
“For networks with limited maintenance windows, this represents a significant operational advantage.”
Smarter planning and better asset management
Unlike traditional condition monitoring systems focused primarily on void detection, KONUX Switch integrates vibration, displacement and shock loading analytics to deliver a more comprehensive view of turnout condition. It enables operators to detect up to 80 per cent of broken crossings at least 28 days in advance, identify 78 per cent of closed crossing defects ahead of failure, and predict 65 per cent of turnout-related delay risks within the same timeframe.
The system also validates maintenance outcomes, identifying unsuccessful interventions and supporting reliability-centred maintenance strategies.
“This shift toward predictive and validated maintenance is helping operators improve reliability while reducing lifecycle cost,” said Weingarth.
Deployment in Australia
Predictive turnout monitoring is now being deployed across Australian rail networks, with Salix providing local engineering expertise, rollout capability and integration with maintenance programs.
Fulford noted that predictive maintenance represents a step change in how rail networks manage turnout performance.
“Smarter, data-driven interventions help cut unnecessary maintenance and extend component life, lowering costs for operators,” he said.
“Just as importantly, reducing emergency callouts allows more work to be planned and undertaken during safer, controlled maintenance windows.
“As rail networks continue to grow and access windows become more constrained, predictive monitoring is becoming increasingly valuable for asset managers and operators.”
A core component of asset management
Fulford said that with increasing network utilisation, improving turnout performance is becoming critical to overall reliability.
“Predictive monitoring is enabling operators to achieve more reliable turnouts, reduce time on track, lower maintenance costs, extend asset life, and enhance overall network performance.
“With deployments already delivering measurable benefits across major global rail networks, predictive turnout monitoring is rapidly becoming a core component of modern rail asset management.”
Through its partnership with KONUX, Salix is working with rail operators to deploy trial programs and develop rollout strategies tailored to network requirements.
Fulford added: “As rail networks continue to evolve, AI-driven predictive maintenance is helping operators improve reliability, optimise maintenance, and extend asset life, delivering tangible operational and financial benefits across the network.”
.jpg){kind=spacer}
