Explanation Page Travel Time Reliability Indicator
A travel time reliability indicator (RBI) is used to represent the reliability of the median travel time over a specific time span of a route. RBI’s are used for policy monitoring purposes for road managers and are not used for real-time traffic information for traffic participants, such as drivers or cyclists.
Implementation of RBI Algorithm in Dexter
Currently, the integrated RBI algorithm only works with the FCD flow. There are plans to extend this to other data flows in the future.
Calculation of the Travel Time Reliability Indicator
The Planning Time Index (PTI) has been chosen for calculating the RBI. This was chosen for the following reasons:
- PTI is not (too) sensitive to outliers
- PTI is easy to explain
- PTI aligns well with practical use
The calculation of the PTI is as follows:
Planning Time Index = 90th percentile travel time / median travel time
The median is chosen because it is not influenced by outliers compared to the average.
Use of the Planning Time Index
The value of the planning time index represents the travel time a traveler needs to arrive at their destination on time with at least a 90% chance.
Example
If the median travel time on a specific route is 30 seconds and the 90th percentile of the travel time is 40 seconds, then the planning time index is 1.33 (40 seconds / 30 seconds). Therefore, if the traveler adds 33% to the median travel time, there is a 90% chance they will arrive on time.
Data Quality Control
Sometimes there are no measurements at certain times. In these cases, travel times or speeds are unavailable. This is taken into account to still report a significant RBI.
It would also provide a misleading view if an RBI were heavily influenced by travel times that are much higher than the allowed average speed. Therefore, it is advisable to cap extremely high speeds. The current FCD provider caps all vehicle speeds above a certain threshold, but this depends on the FCD provider. If NDW switches to another provider, this speed capping is not guaranteed. Therefore, it has been decided to filter out all average speeds above 150 km/h.
Use of Aggregation Periods
RBI per Aggregation Period
For quarterly, hourly, and daily aggregation periods, you will receive an RBI for each aggregation period. For example, if you want to calculate an RBI from 9:00 to 10:00, you will get an RBI based on the entire hour from your selected date range. If you want an RBI for every Monday in a date range, you will receive an RBI based on all Mondays within the chosen date range.
Example
If you request an RBI report with quarter-hour aggregates, you can expect RBIs for each quarter-hour aggregation period:
| route | start date | end date | start time | end time | significance level | 90th percentile | median | RBI |
|---|---|---|---|---|---|---|---|---|
| example route | 03-05-2021 | 23-06-2021 | 00:00 | 00:15 | 0 | 5.01 | 4.69 | 1.07 |
| example route | 03-05-2021 | 23-06-2021 | 00:15 | 00:30 | 0 | 5.09 | 4.68 | 1.09 |
| example route | 03-05-2021 | 23-06-2021 | 00:30 | 00:45 | 0 | 5.07 | 4.63 | 1.1 |
RBI per Week or Month Aggregation Period
For week or month aggregation periods, you will receive an RBI for each week or month. For example, you want to calculate an RBI for week 10 or for the month of August.
Difference with Quarter, Hour, and Day Aggregation Periods
For a week or month aggregation period, you will receive an RBI for a specific week or month, not based on a selected date range.
For quarter, hour, and day aggregation periods, you will receive an RBI over the selected date range.
Example
If you request an RBI report with month aggregates, you can expect RBIs per month:
| route | start date | end date | start time | end time | significance level | 90th percentile | median | RBI |
|---|---|---|---|---|---|---|---|---|
| example route | 01-04-2021 | 01-05-2021 | 06:00 | 07:59 | 2 | 5.46 | 5.13 | 1.06 |
| example route | 01-05-2021 | 01-06-2021 | 06:00 | 07:59 | 2 | 5.46 | 5.11 | 1.07 |
| example route | 01-06-2021 | 01-07-2021 | 06:00 | 07:59 | 1 | 5.6 | 5.2 | 1.08 |
FCD Routes Drawing
You can request the RBI per route and have the option to define your own routes from A to B. You can do this from the RBI report page by clicking the "Select Route" button. Once you press the button, a screen opens within the webpage where you can draw a new route or select from predefined or self-defined routes.
Steps to Draw an FCD Route
- Step 1: Click on a road to define your starting point
- Step 2: Click on a road to define your endpoint
- Step 3 (optional): Click on roads to define intermediate locations
- Step 4: Name your route
- Step 5: Click "save"
If you want to adjust one of your chosen location points, you can do so by moving the positioned points, which will adjust the location.
RBI Significance Level
A significance level is determined by the number of samples available for calculating an RBI. The significance level is an advisory and cannot be seen as a hard boundary. The following thresholds are used to determine the levels:
- 0: Significance level is below the advised minimum (= 50 samples or fewer)
- 1: Significance level meets the advised minimum (= more than 50 samples)
- 2: Significance level is good (= more than 150 samples)
- 3: Significance level is excellent (= more than 250 samples)
Meaning of Samples
The meaning of the samples depends on the chosen aggregation period.
Meaning of Samples for RBI per Aggregation Period
For quarter, hour, and day aggregation periods, you will receive an estimated significance level based on the number of days included in the date range. Each day is considered a sample.
Meaning of Samples for RBI per Week or Month Aggregation Period
For week and month aggregation periods, you will receive an estimated significance level based on the number of quarter periods included per week or month. Each quarter period in a week or month is considered a sample.