As connected-car growth coincides with an expanding conversation about data, the value attached to the information collected from connected vehicles is attracting more and more interest. Over the past few years, we've worked with multiple automakers and fleets to accumulate 18 million cars on our platform. Through this experience, we've gained insights and knowledge about the value of car data that are worth keeping in mind for all service providers.
When measuring car-data value, we observe and analyze five dimensions. They determine:
- The reshaping and analysis process the data goes through in a connected-car data platform.
- The type and location of driver services the data will be valuable for.
- Integration requirements and specifications with automaker and provider systems.
Here is a look at each of the dimensions.
1. Cleanliness: Data cleanliness means how readable, accurate and unified the data is when it is collected. This is crucial for providing correct information to app developers and service providers (e.g., smart mobility, mapping services, maintenance), which can in turn offer trustworthy services to drivers.
For example, in the case of usage-based insurance, if incorrect data from an odometer or driving-style information is provided to an insurance company, the motorist might be charged more and receive stricter penalties. This is a big issue for the driver, the automaker and the insurance company and can have ethical and legal repercussions.
When incomplete data reaches a platform, it should be cleaned and made usable. Sometimes, discovering these issues is not trivial. We might find cars traveling across bodies of water (if they are being moved by ferry), cars seemingly jumping between countries (if they drive through tunnels and lose their GPS signal), recognizing ignition-off alerts when the driving speeds are actually fast and many other scenarios.
2. Freshness: Data freshness covers how far from real-time the data is and is affected by metrics such as transmission delay, refresh rate and latency. Some use cases require near-real-time data (tens of seconds) or high refresh and sampling rates. Others don't depend highly on latency.
Automakers vary in the speed and resolution with which they extract data — from every 30 seconds to once a day. The companies also vary in the speed at which the data travels via their backbone systems, from single-digit seconds to every couple of minutes. Finally, data ingestion also varies among automakers: It could occur every second, every hour or only upon specific occasions such as ignition on/off. All of these variations affect which data can be used and its market value.
3. Location: Just as with real estate, car data usage and value is highly affected by the location of the vehicle. Some countries, states and neighborhoods have higher data demand than others. By analyzing car data in neighborhoods near airports, businesses can plan the best hours to open, understand the most efficient staffing requirements and identify potential rest or leisure areas such as lodging, refreshments, transportation and more.
4. Richness: The richness of the data, also known as the number and depth of the attributes transmitted, also affects the value and potential use cases and services that can be enabled using the data. This determines the value that can be delivered to the end customer. The rule of thumb is the more data, the merrier. But some parameters are more sought after than others — and the complexity of the parameters is crucial. For example, a two-figure GPS measurement is less valuable than a five-figure GPS measurement.
5. Usability: There are many regulations that affect data usability, and rightfully so. After all, we are dealing with people's lives and everyday patterns. Regional regulations exist and include notable frameworks such as the European Union's General Data Protection Regulation, Japan's Act on the Protection of Personal Information and the forthcoming California Consumer Privacy Act. And almost every automaker has its own data governance regulations and internal policies.
These regulations are important to ensure data and users are treated with transparency and respect. But they also affect potential use cases for the data and the applications and services that are empowered. For example, the usage of some data might require driver consent, or advanced systems to implement the right to be forgotten.