In an earlier post, we discussed the potential ownership models for autonomous vehicles, also known as driverless cars (“AVs”). Models range from true traditional ownership as we understand it today, to licensed-based models (vehicles owned by someone else but you can use them on an exclusive or non-exclusive basis), to service-based models (you do not own the vehicle, but you can call it when you want it, e.g. cab, Uber). In this post we will explore the data-intensiveness of autonomous vehicles, the impending data “land grab,” and who will own and control all of the data generated by AVs.
An AV can be thought of as a massive, always-on computer. Sensors in the AV interface with sensors in the environment. Data from satellite navigation systems is also in play. Who owns all that data? Is it the owner of the AV, the ‘driver’ of the AV (e.g. the licensee of a leased AV), or the party collecting the data?
To give context to the question, it has been suggested that AVs will generate about 1 gigabyte of data per second. Consider all the available information – how fast is the AV going? Is the AC on? If so, what temperature? Was that a pothole in the road? The navigation system says turn right at the next light, etc. All of this data is captured and stored – by somebody. That adds up to approximately 1 petabyte of data a year. A gigabyte is 1,000,000,000 bytes. A petabyte is 1,000,000,000,000,000 bytes. For further context, a website reported in 2012 that all the data in the US academic research libraries adds up to about 2 petabytes. So, the quantity of data generated by 1 AV in 30 seconds would max out your cell phone, and the data generated from one AV in two years will match that of all research libraries. The sheer volume is a lot to get one’s head around.
Where will all this data come from? In addition to sensors in cars, there are embedded computers, GPS receivers, and Internet connectivity in the car. There are also sensors on roads, bridges, traffic lights, etc. that are being built into the infrastructure as part of many “smart cities” initiatives. Also, consider that companies, such as Mojio, are developing connected car platforms that will work to gather contextual, behavioral, and diagnostic data on AVs that will be shared in numerous ways as part of the AV ecosystem. So, your travel patterns and behaviors will be collected and used by multiple organizations in a cross-industry fashion for purposes that you may not have contemplated.
Consider SAP’s Vehicle Network, which (according to SAP) was launched in late 2015 to accelerate cloud-based vehicle-centric services. That network includes such disparate players as Hertz, Nokia, Concur Technologies, Samsung, and others to work on technological innovation. Varied players all with different potential uses for data generated by AVs in general or your AV specifically, depending on the application.
With regard to data collection, storage and processing, consent will continue to be the mother of all cure-alls and safe harbors. Recently, I took my Infiniti in for service and received a loaner car. When I turned the car on, the message on the screen advised me as follows:
‘Your vehicle wirelessly transmits recorded vehicle data to Infiniti per subscription agreement for various purposes….’
It then advised that I could limit the transmission of certain data by altering certain settings. So, we’re headed towards an era – indeed, already in one, apparently – in which car owners (or users) will consent to data collection as part of the buyer or user experience.
There are levels of autonomy in vehicles, and different industry players will want different data at various points on the autonomy continuum. As we move from autonomous enablement (e.g. cars that parallel park without driver assistance) towards full autonomy in vehicles such that the person sitting in the classic “drivers” seat has no role to play, the data generated will continue to be varied and will grow, and the organizations that want the data will also be varied and will continue to grow. The data created by your use of autonomous technologies will most likely and increasingly be governed by your consent, or perhaps a lot of the data isn’t ever really yours anyway. The argument will be that the data may be about your travel habits (whether you are the owner or merely the licensee), but it’s not your data. It is owned by its collectors.
All of this collection is not in and of itself inherently bad or problematic, but without guidance there is opportunity for abuse and misuse. There will continue to be opportunities for policy to be promulgated to provide fences for use and to minimize the likelihood of misuse of your personal AV usage data.