Home Mapping solutions & IT Contact

Digi­­ti­za­­tion and the rail indus­try: Is a single frame­work the future?

By La rédaction - 12 February 2019
Reading time: 8min

Today’s frag­men­ted rail­road frame­work siloed by profes­sion raises seve­ral issues: On the one hand, a frag­men­ted frame­work meets the needs of players on-site but does not give infra­struc­ture mana­gers an overar­ching vision of their assets. On the other, a frame­work split into silos restricts itself to data­bases desi­gned for use by only one rail depart­ment (e.g. main­­te­­nance or engi­nee­ring). Without pooled data, the indus­try is fated to a life of useless data dupli­ca­tion and erro­neous legacy updates. The solu­­tion: a unified rail­road frame­work.


As things are today, siloed frame­works are preven­ting updates from being properly deployed, as frag­men­ted frame­works provide know­ledge of legacy equip­ment only at a local level. To increase quality and promote real-time data updates, a unified rail­road frame­work is needed. Yet and still, digi­ti­za­tion in the rail indus­try can only take place if field-level sensors, digi­tal simu­la­tion, and predic­tive main­te­nance are imple­men­ted. For such features to exist, an exhaus­tive, up-to-date rail­road data frame­work must first come to the fore.

Anatomy of a rail network: Why and how to unify it?


Unifying the frame­work doesn’t neces­sa­rily mean buil­ding a unique, centra­li­zed, mono­li­thic system desi­gned as an infra­struc­ture data vacuum. On the contrary, such a system would quickly become diffi­cult to main­tain or modify and consu­mers would lose inter­est in it.

Defi­ning a unified rail­road frame­work means crea­ting a single source of truth—or repo­si­to­ry—­for each data type. It also means forbid­ding data dupli­ca­tion, promo­ting refe­rences, and requi­ring each object to have a unique iden­ti­fier. By relying on these prin­ciples, the fog seems to lift on a unified rail frame­work running on a distri­bu­ted, agile system.


In the rail sector, the goal of digi­ti­za­tion is to increase the added value of rail infra­struc­tures. For this chal­lenge to be met, a solid frame­work is neces­sary before use cases can be massi­vely crea­ted. Rail digi­ti­za­tion makes a number of promises: self-driving trains, more effi­cient preven­tive main­te­nance, and a Pan-Euro­pean train control system.


Self-driving trains

For self-driving trains to operate, they must have infor­ma­tion about the infra­struc­ture they’re opera­ting on. This includes what devices they will inter­act with, physi­cal charac­te­ris­tics (slopes and cants) and opera­tio­nal notions on the track (speed limits), not to mention all the other devices that cross their paths (beacons, signals, rail­road cros­sings). Maps must be exhaus­tive, shared, and upda­ted in real-time. Without a unified data repo­si­tory, the auto­ma­tic train is merely a fantasy.


Preven­tive main­te­nance

Arti­fi­cial intel­li­gence analyzes data streams, cros­sing measures taken by trains with recor­ded cate­nary mecha­ni­cal stress to find corre­la­tions between certain events and break­downs. So long as the rail frame­work is not unified, it will be impos­sible to draw conclu­sions from such measures, which brings to light yet another problem: Why invest massi­vely in data sensors if the bene­fits are non-existent or can’t serve the commu­nity?


The Euro­­pean Train Control System (ETCS)

The stan­­dard defi­ned by the UIC (Inter­na­tio­nal Union of Rail­ways) is deployed by all infra­struc­ture mana­gers. It will even­tually replace fixed commu­ni­ca­tion with train-to-ground commu­ni­ca­tion. Once this occurs, the UIC hopes to stan­­dar­­di­ze train traf­fic at a Euro­pean level and improve line capa­city by conci­lia­ting trains.

To adequa­tely deploy ETCS, the first step is to up the quality of frame­work and, consequently, the preci­sion of sources (drones, LIDAR, etc).


From an infor­ma­tion systems stand­point, a unified rail frame­work is nothing to scoff at, since infra­struc­ture mana­gers will be relying on the stan­dards. Once imple­men­ted by soft­ware publi­shers and compa­nies, such stan­dards will allow costs to be pooled, redu­cing the inte­gra­tion costs of a centra­li­zed data repo­si­tory. Eulynx, RailTo­­poMo­­del, and IFC Rail are just a few of the unified rail­road frame­work adop­ted by GiSmart­­ware to meet the needs of train network mana­gers. To learn more about these stan­­dards, take a peek at our inter­view with Fabrice Simo­­nin, Head of the Rail Divi­sion at GiSmart­­ware!


By La rédaction - 12 February 2019

Leave a Reply

Your email address will not be published. Required fields are marked *

Anatomy of a Rail Network: How and why to unify?
Telecoms GIS: Essential functions for today and tomorrow

Recommended to you

Subscribe to our newsletter

And stay up to date with all the latest news