Case Studies
Turning Rail Operations Data into Actionable Intelligence
SMRT Trains JARVIS Knowledge Platform, Singapore
Critical knowledge fragmented across dashboards, systems, and individuals. Decision-making slowed by the effort of finding, reconciling, and acting on inconsistent data. Maintenance risk rising as institutional knowledge remained locked in spreadsheets and with individual staff rather than accessible to the teams who needed it.
STRIDES Technologies developed JARVIS to solve this at its root, creating a single integrated platform that converts fragmented operational data into consistent, actionable intelligence across the entire SMRT network.
The Challenge
Operational knowledge in a large rail network accumulates across many systems, teams, and individuals over decades. Without a unified way to access and act on that knowledge, decision-making becomes slower, less consistent, and increasingly dependent on individual expertise that is difficult to transfer or retain.
For SMRT Trains, the challenge was not a shortage of data. It was the inability to connect it. Critical information existed across multiple dashboards, maintenance records, and analytical tools, but no single environment brought it together in a way that supported fast, consistent, and informed decisions across reliability engineering, maintenance, and asset management.
The Approach
STRIDES Technologies developed JARVIS as an integrated knowledge and decision support platform, built on Oracle Autonomous AI Database as its core data infrastructure and powered by Oracle Cloud Infrastructure Enterprise AI. The platform unifies maintenance and operations data into a single, trusted source of truth, consolidating train performance metrics, sensor readings, and asset lifecycle information across the network.
The platform connects four core capabilities: reliability engineering, data-driven maintenance, decision support, and asset management. Together, these give teams the ability to identify risks earlier, prioritise actions with greater confidence, optimise resource allocation, and make better-informed lifecycle decisions.
Predictive analytics and machine learning capabilities shifted the maintenance model from reactive problem-solving to proactive intervention, surfacing early indicators of asset degradation before they develop into failures. A natural-language generative AI chatbot interface, enabled by OCI Enterprise AI and vector search capabilities within Oracle Autonomous AI Database, allows maintenance teams to query operational data and retrieve engineering insights in plain language, reducing the time and expertise required to extract value from complex datasets.
The Oracle AI Customer Excellence Centre supported the development, testing, and validation of the platform, providing cloud and AI infrastructure expertise. Oracle Consulting served as implementation partner.
Institutional knowledge previously held by individuals or buried in legacy files was structured, indexed, and made accessible through the platform, ensuring experience built over years of operations is preserved and available to all staff consistently.
Measurable Impact
Faster and more consistent decision-making across reliability, maintenance, and asset management teams
Stronger network reliability through proactive maintenance and earlier risk identification
Lower lifecycle cost through better planning, resource optimisation, and asset life extension
Institutional knowledge preserved, structured, and accessible to all staff across the network
JARVIS demonstrates how AI and analytics, built on enterprise-grade cloud infrastructure, can convert fragmented operational data into practical intelligence, giving rail operators the clarity to make faster, better-informed decisions across every layer of the network.
Services
AI and Decision Support Systems / Predictive and Condition-Based Maintenance / Maintenance Digitalisation / Rolling Stock Maintenance and Analytics