Frequently Asked Questions (FAQ)

Are missing hours equivalent to zero traffic?

No. Missing hours indicate no observed traffic data, not zero traffic volume. Missing records should be handled explicitly in downstream analysis.

Are all speed metrics computed from the same observations?

Yes. All speed metrics for a given segment and hour (average, harmonic average, median, percentiles) are derived from the same underlying set of observed samples.

How should I use segmentId and newSegmentId?

These fields are included for compatibility with historical datasets (only pre-2023 data). For all joins and analysis, dsegId should be used as the primary identifier.

Can a road segment appear in multiple deliveries?

Yes. The same dsegId may appear across:

• Multiple days
• Multiple batch deliveries
• Multiple tiles (depending on delivery configuration)

Users should de-duplicate and aggregate data based on dsegId and time as needed.

How should I handle time zones?

All timestamps and hourly aggregations are provided in UTC. Local time conversions should be applied downstream using the timeZone field from the Road Network file if required.

I get a cryptic error when reading the Avro files - what should I do?

The Avro files are gzip-compressed. If your Avro reader expects an uncompressed container, first decompress the file using gzip, or use a reader that supports gzip-wrapped Avro files.

How large are the datasets?

Dataset size depends on geographic coverage (number of level-8 tiles), time range, and road network density. Traffic Stats Batch is intended for large-scale batch analytics, and users should plan for distributed or memory-efficient processing.

Indicative sizing example (sample-based):

In a sample workload, processing one level-8 tile for 7 days resulted in approximately:

• ~17 million rows
• ~5.6 GB of in-memory data

Actual dataset size and memory footprint will vary depending on the selected area, time range, filtering strategy, and analytics tooling. These figures are provided as guidance only to support infrastructure planning.

Is the data reproducible if re-requested?

Data reproducibility is only partial. This limitation arises because, in accordance with privacy regulations governing GPS data sources, certain data may be removed after several years. As a result, not all historical data remains available indefinitely.

Additionally, the TomTom platform continually expands its portfolio of data sources. This process may include the integration of new sources, even retroactively for past periods. Consequently, historical datasets can change over time as new sources are added.

Can I analyze Traffic Stats Batch data directly in TomTom storage?

No. Traffic Stats Batch data is provided as downloadable files. Customers are responsible for copying the data into their own storage environment and performing all analysis using their own analytics tools and platforms.