deployEMDS

Project details and useful links

What is deployEMDS?

deployEMDS promotes interoperable, trustworthy and accessible data sharing

deployEMDS is a project co-funded by the European Union’s Digital Europe Programme, aiming to bring the European Mobility Data Space (EMDS) to life. The initiative promotes data interoperability, sovereignty and accessibility, encouraging the sharing and reuse of mobility data to support the development of innovative services and applications.

The project includes 16 use cases across nine European countries, supporting real-world implementations in several cities and regions, including Lisbon. These initiatives contribute to improving urban mobility and enabling data-driven policymaking.

TML use case

As part of the project, TML is responsible for developing one of the use cases for the Lisbon region. The title and focus of this use case is Increasing the schedule reliability and/ or commercial speed of buses.

Among the objectives of the software solution is to provide end users with a visual representation of bus traffic disturbances through interactive maps and dashboards.

Concepts

Roadlink: A roadlink is the segment of a public transport route between two consecutive stops of a trip, representing the path traversed by a vehicle from the departure at the upstream stop to the arrival at the downstream stop. It is the fundamental spatial unit used to measure and analyse travel times and their deviations from a reference.

Trip: A trip is a single scheduled journey of a vehicle along a specific route, defined by a direction and a sequence of stops (shape) and an associated timetable. Trips are typically repeated on specific dates according to a predefined service calendar.

Bus Traffic Disturbance Index

Reference - 'Planned'

The planned reference travel time is the scheduled travel time for a roadlink, calculated as the difference between the planned departure times at two consecutive stops. It represents the expected travel time according to the timetable.

For Reference 'Planned', let i=1,…,N index all ride observations for a given roadlink of a trip in the aggregation period (month). Then

Disturbance IndexPlanned=mediani=1,,N(stop_observed_time_end(i)stop_observed_time_start(i))mediani=1,,N(stop_plan_departure_time_end(i)stop_plan_departure_time_start(i))mediani=1,,N(stop_plan_departure_time_end(i)stop_plan_departure_time_start(i))\text{Disturbance Index}_{\text{Planned}} = \frac{ \operatorname{median}_{i=1,\ldots,N} \left( \text{stop\_observed\_time\_end}^{(i)} - \text{stop\_observed\_time\_start}^{(i)} \right) - \operatorname{median}_{i=1,\ldots,N} \left( \text{stop\_plan\_departure\_time\_end}^{(i)} - \text{stop\_plan\_departure\_time\_start}^{(i)} \right) }{ \operatorname{median}_{i=1,\ldots,N} \left( \text{stop\_plan\_departure\_time\_end}^{(i)} - \text{stop\_plan\_departure\_time\_start}^{(i)} \right) }

where:

  • i indexes all ride observations traversing the roadlink during the aggregation period (i = 1, ..., N).
  • median() denotes the median computed over all N ride observations.

Reference - 'Freeflow'

The freeflow reference travel time is the first quartile (25th percentile) of the observed travel times for a roadlink across all rides belonging to the same shape (pattern). It represents a travel time under low-congestion conditions and serves as a benchmark for comparison with observed travel times.

For Reference 'Freeflow', let i=1,…,N index all ride observations for a given roadlink of a trip in the aggregation period (month). Then

Disturbance IndexFreeflow=mediani=1,,N(stop_observed_time_end(i)stop_observed_time_start(i))Q1,shape(stop_observed_time_endstop_observed_time_start)Q1,shape(stop_observed_time_endstop_observed_time_start)\text{Disturbance Index}_{\text{Freeflow}} = \frac{ \operatorname{median}_{i=1,\ldots,N} \left( \text{stop\_observed\_time\_end}^{(i)} - \text{stop\_observed\_time\_start}^{(i)} \right) - Q_{1,\mathrm{shape}} \left( \text{stop\_observed\_time\_end} - \text{stop\_observed\_time\_start} \right) }{ Q_{1,\mathrm{shape}} \left( \text{stop\_observed\_time\_end} - \text{stop\_observed\_time\_start} \right) }

where:

  • i indexes all ride observations traversing the roadlink during the aggregation period (e.g., one month) (i = 1, ..., N).
  • median() denotes the median computed over all N ride observations.
  • Q₁shape() denotes the first quartile (25th percentile) of the observed roadlink durations computed across all rides belonging to the same shape (pattern). This value is used as the Freeflow reference.
  • A shape is defined by the combination of the stop pair and the corresponding shape extension in metres.

Disturbance classification

The Disturbance Index is a relative deviation. The relative deviations in duration (travel time) between stops are then reclassified based on this table:

Disturbance ClassDisturbance IndexInterpretation
Severely EarlyR.D. < -0.20Observed duration is 20% or more below the reference duration.
Early-0.20 ≤ R.D. < -0.05Observed duration is between 5% and 20% below the reference duration.
On Time-0.05 ≤ R.D. ≤ 0.05Observed duration is within ±5% of the reference duration.
Late0.05 < R.D. < 0.20Observed duration is between 5% and 20% above the reference duration.
Severely LateR.D. ≥ 0.20Observed duration is 20% or more above the reference duration.
UnknownN/AInsufficient data to determine the disturbance class.

Disturbance Dashboard

Bus Traffic Disturbance Dashboard is available here).

Disturbance Data

Bus Traffic Disturbance Data can be downloaded through API endpoints documented here.

On this page