AI planning platform streamlines railyard operations and cuts daily planning from hours to minutes
A global railyard operator needed to replace manual scheduling with fast, accurate optimization.
Digitized railyard operations with AI/ML optimization, cutting planning time from hours to minutes and improving operational agility.
Client: A multinational railyard operator relied on manual, paper-based processes for daily planning. This approach slowed decision cycles, introduced inconsistency, and limited visibility across sites.
The company partnered with Lumenalta to build an application that produces optimized intake and routing plans using machine learning. Planners gain clear recommendations they can review and apply within a simple interface.
The system reduces dependence on a small group of experts and shortens training time by capturing and re-using domain logic within the platform.
Challenge: Creating a modern planning platform for complex yard activity required solving several specific issues.
- Manual scheduling consumed most of the workday and varied by site.
- Data lived in separate tools, limiting a single view of railcar movement and capacity.
- Planning logic differed across teams, which led to uneven results.
- Legacy systems did not connect to real-time inputs from the yard.
- A small number of experts carried critical operational knowledge.
- Limited forecasting made it hard to spot bottlenecks or conflicts early.
- On-prem systems struggled with the computing needs of machine learning.
To meet performance goals, the platform needed to produce accurate recommendations, scale across locations, and maintain clear oversight for planners and leadership.
Solution: Lumenalta built a cloud-hosted planning application that analyzes historical data and current yard conditions, then suggests optimized intake and routing plans for each cycle. Recommendations update as conditions change, providing a practical guide for daily work.
The application uses Azure SQL, Azure Service Bus, Python, .NET, Kubernetes, and Azure DevOps—a stack selected for reliability, scalability, and ease of deployment.
A clear interface lets planners adjust inputs, validate outputs, and commit plans. Workflows and documentation support handoffs, reduce reliance on individual experts, and shorten onboarding across sites.
Outcomes: The new system produced measurable gains across time, quality, and scale:
- Daily planning time reduced from 10 hours to 15 minutes at U.S. locations.
- Application rollout completed across ten U.S. sites.
- Recommendations adapt to changing conditions in near real time.
- Lower dependence on a few experts; faster training across teams.
- Standardized planning logic improves consistency across operations.
- Clear stack selection supports reliable deployment and maintenance.
The platform now serves as the daily planning foundation, with consistent outputs and faster cycles across locations.
Reduced
planning time from 10-15 hours
Completed
application rollout across ten U.S. sites
Adaptive
recommendations respond to changing conditions in near real time.