Data seen as an essential cost to attract consumers
To alleviate the cost of Big Data one must first ask how it is defined. Therein lies the problem. Most articles and papers on Big Data for connected and autonomous vehicles prefer to focus on its value because it offers a more positive and promising perspective than any discussions about the physical, wireless and software-related infrastructure that support the use of Big Data in connected car and trucks. For example, a 2014 report by automotive technology specialist SBD entitled Automotive Big Data: What’s It Really Worth? claims that value of Big Data is around “$130 (£98) per car per year based on 15 use cases”.
The report’s authors, Vishnu Muralidharan and his SBD colleague Lee Colman – both specialists in connected cars, admit that “collecting Big Data is expensive [because] the cost of embedding connectivity, the network data costs, data storage and processing costs are significant”. They, nevertheless, believe that the data collection costs can be covered by creating a stack of “multiple use cases on the same platform and data”. This approach will permit carmakers and OEM suppliers to align their sales, CRM, warranty and servicing departments to enable them to exploit the data that they collate from each vehicle for each use case for the customer’s benefit. It is, therefore, generally accepted that Big Data will drive connected car services well into the future.
Increasing memory
Volker Schumann is senior manager, automotive sales, Toshiba Electronics Europe, points out another issue that will affect the cost of Big Data in connected cars in his 1st March 2016 article for Electronic Weekly: The need for “big memory for Big Data”. With Big Data comes the need for faster processors and the need for more memory capacity and he estimates that the latter will expand from 15GB in 2018 to 60GB in 2025. In fact he says 70-80% of all storage is taken up by mapping and media in cars.
“Much of this growth will be met by solid state drives using NAND flash memories.” He also believes that the advance towards advanced driver assistance systems will further expand the need for even more memory intensive multi-layer high definition maps. Voice recognition and speech synthesis applications will ultimately be party to this trend too – particularly as there are a number of cars that already support natural language speech recognition.
Enabling insurers
Big Data has a range of applications. They can be used to enable insurers, for example, to assess the driving risk and risk of a particular person with the use of telematics. Robin Harbage, director at Willis Towers Watson, explains: “Many insurers find that Big Data is already paying its way due to segmentation.” Big Data is enabling them to become more profitable as they become more able to make more appropriate risk assessments and this allows them to offer set better pricing.
He adds that the actual cost of the data storage is not the inhibitor of UBI. “The problem is the cost of the data collection technology: The devices and cellular transmission costs have been prohibitive,” he claims. Yet the cost is coming down because most data is collected through smartphone applications, and this is enabling more mass marketing. He believes the next step is to collect data from any existing embedded technologies in cars or phones to “pave the way for even less expensive and faster adoption because it takes no additional effort on the part of consumers, other than to give insurers permission to use the existing data”.
A need to understand
Magnus Johansson, director business development, WirelessCar thinks that the connected vehicle industry doesn’t fully understand the potential of Big Data and its associated costs yet. He, nevertheless, says that “everyone has some kind of feeling that there is something here, but it’s difficult to get a handle on it and so it’s not achieving its full potential”. To realise this potential he advises them to focus on analysing the value of Big Data.
“Big Data is a way to improve products, reduce warranty cost and call-backs and it’s a good way for the OEM to create loyalty with the customer to make sure they are using the tools and workshops that are preferred by the OEMs,” he says before suggesting that there “may also be a market for selling data to companies that are aggregating traffic information but this is step two.” For now he thinks automakers should focus for on making their own operations better.
Holding back data
Talking of data sharing, which can often be held back by the need for data protection compliance, Harbage says: “OEM’s have become very willing to share driver data, so long as they can obtain permission from their customers. A large hurdle is to normalise the data, then transmit it in a usable fashion to the hundreds of separate insurers who can then put it to use in their pricing. This is the problem Willis Towers Watson is solving.”
Johansson describes data sharing as a “delicate situation [because] privacy and integrity mean that there is some resistance to opening up access to personal data”. Automakers and their connect car ecosystem partners need to first of all, as Harbage indicates, ensure that they have gained explicit consent from their customers before this can ever take place.
“You have to tell them what you are going to do with their data,” stresses Johannsson who says that the current legislation is making everyone nervous and as a result it is slowing down the opening up of access to the data. He doesn’t feel that there will be much progress with data sharing in the short-term. For the moment automakers, application developers and dealers will want to ensure that this activity doesn’t create any risk to unauthorised third-parties from gaining access to the data.
Applink and Carplay
He adds: “Solutions like Applink and Carplay are examples of how apps can get access to the audio and display of the vehicle but not to the car itself. The OEM’s are unlikely to give too much access to the vehicle to 3rd party developers because of the liabilities attached to it, but there are initiatives to create some form of gateway within the vehicle to allow access to certain information – to an app or to a head unit.” Data sharing is, in his view, likely to be more prevalent in the commercial logistics sector where fleet management systems have the ability to access truck data to monitor drivers, the location of the vehicles, weight and fuel consumption to enable good integration into the logistics system.
“A vital key to sharing data is transparency,” emphasises Harbage. He agrees that the data aggregators need to be sure consumers are told what data is being used, how it will be used and with whom it will be shared. “If consumers are told this in clear, unambiguous language and subsequently asked for permission to use it in a manner that benefits the consumer – then adoption and trust will be achieved,” he claims.
Pre-emptive response
Johannson agrees that Big Data offers value for OEMs as it can enable pre-emptive servicing and targeted advertising. Wider application ecosystems can offset and alleviate the high cost of data management, too, because they can contribute to the total business case for connected cars. OEMs, for example, can use Big Data to discover how the product is used for future product and service development to reduce warranty costs and improve customer loyalty.
He also believes that there are opportunities to sell vehicle and driver data, providing it’s done in the right way. “You can detect, for example, road conditions which will have a certain value and there is a possibility for OEMs to create packages for finance and insurance,” he explains. Yet, as the industry develops autonomous vehicles, there will be a need for more robust and secure connectivity for the download of data from all of a vehicle’s sensors and on board cameras, as well as updated software. He concludes that “this would put pressure on bandwidth and someone has to take up a role in the middle to standardise the data, and this is a challenge that is common within the industry”.
Therefore, there is a constant need to understand and address the costs and the value of Big Data, and as connected vehicles and autonomous vehicles progress this pressure is bound to increase. The problem is that nobody seems to be addressing what the costs are but it appears that the best way to alleviate the cost of Big Data management is by constantly seeking new ways to add and offer value.
