Introduction:
Mooven can handle a wide range of datasets. The way this information is presented and updated depends on the dataset and use case.
Common variables include:
Individual record data vs aggregated data
Frequency of updates
Calculated metrics, for example, average speed or 85th percentile values.
This article provides additional information on data collection logic, focusing primarily on traffic data.
Individual Vehicle Data
In these scenarios, a stream of events is captured for each vehicle or record captured. Common examples include live bus performance monitoring and Automated Number Plate Recognition (ANPR) monitoring as shown in the two figures below.
Bus Journey Time Example:
In the above example, each icon represents a completed bus journey between the defined segments, with the datapoint recorded against the completion time (i.e. when the bus completed the journey).
Within these visualisations, users can add an aggregation layer using the controls in the top left-hand corner to view the maximum, minimum, 85th percentile or median value. These calculations are based on 15-minute time bins.
The update frequency of individual record data is dependent on the source. Three common examples include:
Bus data - updated every 30-60 seconds.
ANPR cameras - the camera pushes completed records to Mooven in batches every 30 seconds to two minutes when there are completed trips.
Event and Incident data - updated every 15 minutes.
ANPR Camera Example:
In the above example, you can see the completed trip time for vehicles between two ANPR cameras. Each dot represents a completed individual journey.
This example was used to demonstrate the impact of stop-slow traffic control, which was active on this date. As the vehicles were temporarily stopped for up to 10-15 minutes, you can see vertical stacking of the results as groups of vehicles completed their journey around the same time.
The above visualisation clearly shows how individual vehicles may have quite different experiences from each other and the average experience. The time to complete the journey, if you were the first or last car to get stopped, ranged from trips in ~25-30 minutes to trips lasting up to 55 minutes. While a complaint about a 55-minute journey in this example would be real, it does not reflect what most vehicles experience.
Aggregated Vehicle Data:
By contrast, many providers use data about all current trips to provide accurate information on the expected journey time for anyone who is travelling now.
Floating Vehicle Data:
Mooven uses floating vehicle data from providers like TomTom, Google and HERE to source rich, live traffic information.
Google is typically the default source used due to it's market share and ability to provide live data on a larger network. Google analyses the GPS, cellular, and Wi-Fi locations transmitted to it by mobile devices across Android and iOS platforms, in addition to connected vehicle data (where Google is used within car navigation) and third-party sources to provide accurate current live travel times. Google processes the incoming raw data about mobile phone device locations, then excludes anomalies and removes all personally identifiable information. TomTom and HERE both operate similarly.
Mooven can poll floating vehicle providers' services for the current journey times for a defined route. This is the same process you would use if putting directions into your car or mobile phone, but in a way that enables Mooven to make millions of requests programmatically.
Each provider uses its proprietary algorithms to exclude anomalies and outlier data that could negatively impact an accurate prediction (e.g. a garbage truck moving slowly down a road that is not congested)
With floating vehicle data, the request represents the live journey time if you started the journey now. That is, values are timestamped to the start of a journey rather than the end of a journey.
One advantage of floating vehicle data is that you don't need vehicles to complete the forward journey to get data on escalating traffic congestion, unlike many hardware sources that won't record the issue until the vehicle completes the journey, resulting in delayed notifications for incidents. Conversely, floating vehicle data can under-estimate the extremes of congestion, as it either won't show you 85th percentile travel times or may filter vehicles that are stationary for an extended amount of time out of the report, as it can't determine if they have stopped for another purpose (i.e. getting a drive-thru coffee)
Floating Vehicle Data Example
How data is collected:
Mooven requests the latest current journey time on the route defined.
A request is made every 15 minutes for normal monitoring or every 5 minutes for queue and variable message sign monitoring.
As providers limit the number of requests sent per second, Mooven balances requests across all customers over 15-minute time windows. The polling of individual routes within a site are syconised to make sure that results can be accurately compared with each other. For example, a specific site might be set up to update at three minutes past the hour and on a 15 minute update schedule: 6:03 pm, 6:18 pm, 6:33 pm and 6:48 pm.
The results captured are shown against the closed 15-minute time block. For example, any results after 5:52:30 or before 6:07:29 will be associated with the 6:00 am time block.
If multiple results are captured in the timeblock, for example with queue monitoring, the maximum (worst case) result is displayed.
Volume Summary Example:
In the above example, traffic volumes have been aggregated into hour time blocks regardless of whether the underlying data includes vehicle-level information or if aggregated information was provided in the first place.
This level of aggregation is common for datasets like vehicle count. Depending on the source, Mooven will also show a time-series of volume within 15 minutes time bin using the same logic as described for journey time and delay data.
Event and Incident Data:
Event and Incident data collected from TomTom and HERE is updated every 15 minutes by default. Events and incidents are individual events and as such will be recorded against the event time.
Direct Integrations:
Customers can establish direct integrations for both traffic data and also the sharing of project information or other feeds like events and weather. In these instances the update frequency is specific to the requirements, however the general behaviour of data displayed will match the patterns described above.
Data collection Faults
If Mooven is unable to collect the planned data point(s), our data pipelines will retry automatically, which resolves temporary errors without missed data points. If we detect an ongoing failure to connect data, the appropriate users will be notified via email. See our article on polling failure notifications: https://help.mooven.com/en/articles/9257827-handling-data-collection-issues-on-your-site