The Vehicle Driving Events dataset delivers classified acceleration, braking, turning, and eco-driving events from smartphone-based telematics across the USA, complete with motion magnitude, directional angles, and speed measurements at the event level. For the first time, your UBI models can price on observed driving behavior instead of demographic proxies, without building a data collection pipeline to get there.

Most data vendors sell you a national license to millions of records you will never contact. BigGeo works differently.You tell us where you sell, and we cut you exactly that geographic slice of the dataset. A single city.
A cluster of ZIP codes. A metro area. A county. Whatever matches your territory. The result is a lean, CRM-ready file containing only the companies in the markets you actually work, delivered at a fraction of the cost of a full national license.

Auto insurers know that zip code, credit score, and vehicle age are weak proxies for how someone actually drives. The UBI promise has always been pricing on real behavior. The gap between that promise and what most teams can actually execute on comes down to one thing: access to clean, structured, event-level telematics data at scale. Without it, your risk models are still guessing.
Every record is pre-classified as acceleration, braking, turning, or eco-driving. Skip the signal processing work and go directly to building frequency-per-mile risk features your actuarial team can use today.

Speed measurements at the beginning and end of each event let you calculate velocity delta, a core variable for distinguishing aggressive maneuvers from normal driving and calibrating the magnitude of surcharge triggers.

Turning events include the directional angle of the maneuver, giving your model the ability to separate sharp, high-risk cornering from gradual turns and assign differentiated risk weight accordingly.

Start and end coordinates on every event mean you can join driving behavior to road type, speed limit, intersection density, or weather exposure, unlocking the geographic risk layering that flat telematics files cannot support.

BigGeo AI is live in ChatGPT today and shipping in Claude, giving your team direct access to governed driving event data through the tools they already use. Instead of waiting on a GIS analyst to pull event clusters by region, your actuary or product manager can ask which corridors concentrate the highest hard-braking frequency and get a spatially grounded answer in seconds, backed by real 2025 data, not a model hallucination. No GIS software. No data pipeline. No waiting.
The dataset is provided by Driver Technologies, a company whose core business is smartphone telematics infrastructure built for commercial deployment. Data collection through their platform operates within the consent and privacy frameworks required for commercial telematics use cases. If your legal or compliance team needs additional documentation, that is a conversation we can facilitate through the sample and access process.
The dataset spans continuous 2025 collection across the USA with events attributed to individual trips, giving your team the raw frequency needed to compute per-driver metrics like hard-brake rate per mile or turn severity distribution. Once the data is in BigGeo's DataLab, you can join it with your own policyholder records and filter for minimum event thresholds before any score is computed.
The dataset reflects 2025 collection from live smartphone-based telematics, making it among the most current behavioral driving data commercially available. Coverage and record volume reflect ongoing collection through the year. Speak with us about refresh cadence and whether a scheduled update fits your model retraining cycle.
Once purchased through the BigGeo Marketplace, the dataset lands directly in your DataLab workspace where it can be queried, joined to your own data, and exported in formats your engineering team already works with. There is no proprietary SDK to adopt and no separate ETL pipeline to build. Velocity, the compute engine underneath DataLab, executes spatial queries at sub-second speed so your data science team is not waiting on batch jobs.
Request a sample on this page and we will send you a representative slice of the driving events dataset scoped to a geography relevant to your book of business. From there, a 30-minute call covers what full access looks like, how the data maps to your existing risk model inputs, and what a pilot would take to stand up. Most technical teams are running their first queries within a week of access.