AskRail®
Read More ›
TECHNOLOGY FOCUS
Building on the work we have done in past years, Railinc is continuing the use of artificial intelligence/machine learning (AI/ML) solutions for our customers.
Recently, our Data Science team has made use of density-based spatial clustering of applications with noise (DBSCAN), a machine learning clustering technique, in committee work related to wheel temperature data.
2022 Product Updates
As part of Railinc’s mission, we continuously seek to develop our products to serve the changing needs of our customers and our industry.
This year we introduced several updates focused on these areas in our applications:
2022 Project Highlights
Railinc Employees Tackle Data Migration
One of our biggest projects was completed in 2022: the cloud migration.
As of June 2022, Railinc successfully migrated more than thirty Railinc applications and their modules to the cloud, marking the end of a project that took over two and a half years to complete. Many Railinc employees had a direct hand in this project and spent late nights, weekends and countless hours making sure this would be successful.
The project was approved in 2019 and launched in 2020. Despite the workforce shifting to a remote and hybrid work model during much of the project, the entire cloud migration was completed ahead of schedule thanks to the dedication of our employees.
Throughout the project, the cloud team found support from their managers and their Railinc peers. When any issues were identified, management worked toward a solution, and the cloud migration team coordinated with product support to go over the intricacies of each application. It is this sort of collaboration that makes Railinc one of the top places to work in the Triangle – and a frequent talking point of our team members.
Data Science projects
Improvements to ETA Service
Railinc’s AI-enabled advanced ETA service was bolstered by continued enhancements in 2022. One of the biggest impacts to the service was the introduction of a new pre-training method where Railinc achieved an improvement significantly higher than the previous implementation in both training time and cost. This enabled Railinc to double the number of training lanes – from 1,500 to 3,000 – and reach new customers interested in improved ETAs. In addition to this, a wrapper mapping service was also created to capture additional lane traffic, enabling Railinc to extend coverage by twenty percent.
Railinc Takes Part in Proof-of-Concept GPS Pilot
GPS data has the potential to transform rail industry services, including improved Estimated Time of Interchange (ETI), better railyard management and car pinpointing, and demurrage calculations/dispute management. Railinc and several Class I railroads collaborated in a GPS proof-of-concept pilot expansion in 2022, reaching 1,000 installed devices with up to 18 months of historical data. In addition to advisement on data cleaning and processing, Railinc devised an experiment to improve ETIs using geofences and GPS ping data. By aggregating GPS pings according to origin-destination pairs, and constructing concentric geofences around the destination, Railinc showed improvements in initial predictive estimated time of arrival assessments. As the project progresses, Railinc expects better location reporting which would lead to logistical improvements for railroads and higher customer satisfaction, as customers would have better visibility of the location of their cars.
Looking Ahead: 2023 AAR Industry Projects
Much of our technology project work originates each year in committees sponsored by the Association of American Railroads (AAR). These committees work to ensure the smooth and efficient interline movement of freight, identifying common business problems that can be addressed through centralized data and technology resources. They give industry stakeholders the chance to provide input into rules, processes, and technologies that contribute to the safe and efficient movement of freight by rail in North America.
We’ll be working on 21 approved projects throughout 2023. The full list can be found on our AAR Industry Projects page and we’ve included a few below.
For our Interline Services, the AskRail® application will be further enhanced to allow training for first responders and will increase the training materials and scenarios available.
Finally, in the Positive Train Control (PTC) system, our team is building on work done in previous phases of our Fully Scalable Train Braking Simulation Environment. The project seeks to host, execute, store and report on Braking Algorithm scenarios executed by MxV Rail, which MxV Rail will then use to improve efficiencies and expand the capabilities of the current PTC braking simulation application.