Telemachus

Telemachus is an open, Parquet-native data format designed specifically for high-frequency telematics data. It addresses the challenge of unifying scientific kinematics data—such as multi-rate GNSS and IMU readings at 10-100 Hz, including accelerometer gravity frame tracking—with scalable fleet analytics that incorporate OBD, trip metadata, and carrier state information. The format is intended to be instantly queryable in data analysis tools like Pandas, Spark, DuckDB, and Athena, facilitating efficient data handling for both scientific and fleet management applications.

A key feature of Telemachus is its vendor-agnostic schema, which supports a range of data types including GNSS, IMU, OBD, I/O, and CAN. The schema is profile-based, offering configurations for core GPS data, IMU with accelerometer, and a full profile including gyroscope data. yaml file to support reproducibility. The specification is peer-reviewed and organized around four main pillars, as outlined in the normative spec.

Telemachus is aimed at data scientists and researchers working with telematics logs who require a consistent schema to avoid rewriting pipelines for different vendors. It is also designed for datalogger designers who need a standardized target format for device logs, complete with a validator and test suite. The platform provides Python tools for reading, validating, converting, and introspecting data, and includes reference adapters for open datasets such as AEGIS, PVS, and STRIDE. The toolkit can be installed via pip, and there is a demo notebook available for hands-on exploration of the format.

The project is open source and released under the MIT License, with schemas, reference adapters, and Python tools all available. Telemachus is maintained in a single monorepo on GitHub and is positioned as a pivot standard for mobility and telematics data, distinguishing itself from other standards by providing a columnar, instantly queryable, and reproducible format.