Application and content management framework for railway operators Dr. Bruno Volckaert – Ghent University / iMinds (Belgium)
Large # of applications Unreliable Connectivity
Cause: diverse fleet (manual) software maintenance 12/09/2018 Cause: diverse fleet (manual) software maintenance Train: out of service Train: into depot Engineer: go to train Engineer: file upload from storage Engineer: local test Train: back in service
RAILS framework TRACK RAILS 12/09/2018 Automation in railway application & data management Robust distributed system Software execution environment Software provisioning platform Software lifecycle management Robustness and reliability Legacy integration Built on solid technology Small requirements, low footprint while running
Software execution environment RAILS Platform Wayside server Train node
Software execution environment Apps Core apps & services RAILS
Core applications and services Core apps & services Advanced remote administration Connectivity service Reliable messaging service Persistence service And many more…
Automated provisioning Application Media Data Calculate target devices Calculate target devices Calculate target devices
Administrator overview of provisioning Module store Hot-updating (while driving) Context-aware scheduling Due date Available connection Timetable information
Dealing with limited bandwidth: delta provisioning
Module lifecycle and dependency management
Robust and reliable Server unavailability: auto-relocation of clients Nodes are built for unreliable connections Outgoing data is queued Data sent once connected
Other notable features Legacy applications inherently supported Built on proven technology OSGi, Java, RESTful services, JSON Small resource footprint RAILS base platform installation Node (e.g. train): < 20 MB Module Store: < 20 MB Server : < 25 MB Memory usage while running is < 256MB
Conclusions and future work Scalable and robust software / data management framework for railway operators Research solution integrated in iSync by Televic RAIL, deployed in e.g. Eurostar fleet Calgary (Siemens) Translink (CAF) MUNI light rail San Francisco (Siemens) Current research Extending application / data management towards train sensor layer Rapid application / data flow development through LimeDS (https://limeds.intec.ugent.be)
THANK YOU FOR YOUR TIME Questions? Contact information: bruno.volckaert@intec.ugent.be LinkedIn: https://be.linkedin.com/in/bvolckae