Teradata explains why data privacy is more important to the connected car than you’d think

For car makers, harvesting, storing and securing data from the connected car can get pretty pricey. So the ability to treat data differently based on its sensitivity and value to the organisation is a must. 

Torbjörn Rosenquist, Automotive Industry Practice Leader | Teradata International

With data breaches still grabbing the headlines, there is one area that car makers need to urgently address before we see a significant uptick in connected car adoption – that of data privacy. According to a recent global survey, 51% of German, and 45% of U.S. customers said they had reservations about using connected car services because they were worried about digital safety and data privacy.  The first is the possibility that connected cars could be hacked and sabotaged by nefarious individuals.  And the second is the concern that the connected car data could be shared without the knowledge or consent of the car owner.

The industry has started to address this issue, with the Alliance of Automobile Manufacturers and the Association of Global Automakers putting forward a set of privacy principles for vehicle technologies and services in November 2014.  And a number of car makers, such as BMW and Volvo have also stated publicly that they are taking this matter very seriously, eschewing possible monetization opportunities in favour of putting their customers’ privacy first.

Don’t Underestimate Data from the Connected Car

The connected car has been called a “gigantic data-collection engine” for good reason. Today, car makers might be downloading 100 – 200 kilobytes of data from a car, once a year, during its annual service. But with the connected car, kilobytes of data can be downloaded every day. And then there is also the remote diagnostics capability of recording data on-demand as needed to study anomalies in detail for vehicles when the need occurs.

With Gartner predicting that there will be 250 million connected cars on the road by 2020, the scale of the data challenge for the automotive industry becomes clear. Companies will need to decide whether it makes business sense to collect all the data available, considering the cost of transferring and managing the data.

Still, there are areas that data from the connected car can be game-changing.  Product innovation is now taking place at such speed that a unique innovation today can easily be copied and manufactured by a rival firm soon after.  In short, product specifications are no longer a unique reason for the customer to choose one car model over another.

Connected car data can be invaluable in this instance as it can help car makers understand how to create new services and experiences for their consumers that will differentiate their cars from the competition. Services like allowing car owners to receive shopping deliveries securely to the boot of their car. 

And with industry trends indicating that the nature of car ownership itself is changing, the connected car data can help car makers position themselves to offer alternatives such as pay-as-you-drive insurance, car leasing or shared-usage schemes. 

Defining Privacy

But before deciding how data from the connected vehicle can be used, companies need to first understand what data should be labelled “private”.  Some data can be collected and analyzed in an anonymized format – data that transmits the condition of a vehicle’s function and quality, or driving behavior, for instance.  This data, even when anonymized, could still be of value to car makers (more on that later).

Next, consider geolocation data from a connected car.  Car makers, in conjunction with other companies could use this to provide services to the driver – advertising the nearest petrol station or services for instance. But car makers first have to convince customers that this data is stored securely.  Secondly, car makers will have to ask customers to opt-in to share this data in exchange for the service.  That means convincing their customers that there will be a true value exchange for their data. 

There are already signs that this can be done successfully. Just have a look at vehicle insurance policies that are tied to telematics.  In 2013, in the UK, there were 300,000 such policies, triple the numbers from 2011. A clear indication that customers would be willing to share their data if they feel that they are getting a good deal in return.

Privacy in Practice

For car makers, harvesting, storing and securing data from the connected car can get pretty pricey. So the ability to treat data differently based on its sensitivity and value to the organization is a must. 

That could mean storing data that can be used to identify individual consumers behind the corporate firewall instead of in the cloud to ensure data security.  On the flip side, the second-by-second breakdown of driving data based on position is best kept in the cloud rather than behind the firewall to keep a lid on costs. It can still be accessed by the car maker or a business partner, say, an insurance company, who would use this data to bill for pay-as-you-drive insurance. This data can be kept for the period where it creates value and then deleted.

Not all data has proven value to the organization.  Treating all data equally, regardless of its value to the organization would quickly become expensive. With the unit cost of data storage at historical lows, many companies are already storing big data sets that are infrequently accessed or which are of unproven value in cheaper storage options. 

Getting Value from Anonymized Data

With 50% of car warranty costs now said to be related to electronics and software, there is a market opportunity for car makers to use anonymized data from the connected car to decode engine faults.

Take time series data, for example. Time series data is already being collected from running engines, engine testing and normal certification cycles.  When a failure event takes place, engineers can also collect time series data from the affected vehicle using freeze frames. By comparing the data patterns from freeze-frames collected in failure events, with time series data from normal running engines, and other data around vehicle configuration, market data, service data, and engine software versions, manufacturers can understand not only the root-cause of the problem, but also the population of vehicles that might experience similar issues. They can then decide if this analysis should be done on a continuous basis or only until the corrective action is implemented.  And they can decide if further testing procedures are needed for engines that are currently being manufactured or under development. Powerful stuff, considering the amount car makers put aside to cover warranty costs.

Winning Customer Trust as a Springboard
If customers feel that they can trust car makers with their data, just like they trust a bank with their monetary assets, then the connected car will become mainstream.  But for car makers it is crucial that this discussion of privacy starts with an evaluation of the data.  Not all data is created equal.  Some should be kept more secure than others.  Some have value to the company even when anonymized. Taking the time to build a data strategy that addresses this issue of trust is actually a really good place to start for car makers looking to create value from their connected car data.
Please see the Indian automotive telematics fraternity gathering @ Telematics India 2015