Participate
Share your datasets and applications with the open data community
The historical method used to calculate the statistics presented in the dataset Monthly imputation of delays takes into account all the minutes of delay caused by 'major incidents' (internally know as 'relazen') on the rail network as reported to the Railway Accident and Incident Investigation Body (OEAIF/OOIS) and the Railway Safety and Interoperability Service (NSA Rail Belgium) under the Royal Decree of 16 January 2007 laying down certain rules relating to investigations into railway accidents and incidents. The criteria defining 'major incidents' (internally known as 'relations'**) are as follows: 1 passenger train delayed by an incident for 20 minutes or more Several passenger trains delayed by an incident for at least 40 minutes Incidents leading to the cancellation (partial or total) of trains Incidents with an impact on operational safety There is no unequivocal relationship between the minutes of delay in 'major incidents' and the punctuality rate because: The minutes included in 'major incidents' do not necessarily have an actual impact on punctuality (a train can make up its delay as it goes along). Some trains arrive at their terminus more than 6 minutes late (and therefore have an actual impact on punctuality), but are not included in the 'major incidents'. In order to provide an exhaustive overview of the causes and responsibilities for delays, a new dataset has been made available: Monthly causes of loss of punctuality. The data presented in this new dataset is as follows: For each train delayed by 6 minutes or more on arrival at a tracking point*, an analysis is made of the cause of all the minutes of delay along the route, and a proportional score is awarded for each responsibility identified. Example: IC train 507 Ostend-Eupen leaves Ostend on time, arrives at Bruxelles-Midi two minutes late (signalling problem, Infrabel's responsibility), and is 6 minutes late on arrival at Eupen (a further 4 minutes late due to faulty rolling stock, SNCB's responsibility). The only late sighting of this train was at Eupen station, but the two minutes delay in Brussels are also counted. Imputation Infrabel: 2 minutes out of a total of 6 minutes, so 1/3 of the delay for this late sighting = 0.33 of the late sighting is Infrabel's responsibility. Imputation SNCB: 4 minutes out of a total of 6 minutes, i.e. 2/3 of the delay for this late sighting = 0.66 of the late sighting is borne by SNCB. The late spotting scores added up for each of the parties are then compared with the total number of late sightings. In January 2022, Infrabel was responsible for a total of 1153.5 late sightings. This is equivalent to 1.02% of total train sightings over the same period (113432 sightings). In January 2022, SNCB was responsible for a total of 2556.98 late sightings. This is equivalent to 2.25% of the total number of train sightings over the same period (113432 sightings). In January 2022, train punctuality was 93.6%. The 6.4% of late sightings were attributed to Infrabel (1.02%), SNCB (2.25%) and other stakeholders (the remaining 3.13%). The categories to which delays are assigned are defined as follows: Infrabel: delays generated by the infrastructure manager in terms of infrastructure reliability, works management and traffic control NMBS/SNCB: delays attributed to NMBS/SNCB in terms of train service operation and rolling stock reliability Third parties: delays due to causes not directly related to the activities of Infrabel and the railway operators. Examples : accidents at level crossings, collisions with persons, cable theft, etc. Note : the 'third parties' category represents a smaller proportion of delays in the 'all minutes' data than in the 'major incidents' data, because major incidents usually lead to the opening of a 'relation'. They are therefore overrepresented in the 'major incidents' data ; the 'all minutes' data gives a more objective picture of the proportion of delays actually caused by third-parties. Systemic robustness: certain causes that are not visible in the analysis of major incidents, but which nevertheless have a major impact on punctuality in reality. Example : a train is delayed 10s in each stop and then has a 2-minute delay in Brussels, causing it to hinder another train. Others: delays generated by other railway operators and delays with no identified causes. *a train is considered to be late when it arrives at a tracking point 6 minutes or more behind schedule. The tracking points are the terminus station, and if the train uses the Brussels North-South Junction - the first station on this Junction that is on its route. **relation : all events constituting an accident or accidents or an incident or incidents involving or having an impact on operational safety that occurred on the railway network (Royal Decree of 16 January 2007 laying down certain rules relating to investigations into railway accidents and incidents)
Share your datasets and applications with the open data community