In the first of a two-part series, Andrew Tolve explores the Big Data opportunity – and why so few automakers have capitalized on it.
The average car on the road today generates a mind-boggling amount of data. With sensors monitoring everything from tire pressure to engine RPM to oil temperature and speed, cars can produce anywhere from 5 to 250 gigabytes of data an hour.
Advanced concept cars go even higher; Google’s autonomous vehicle, for example, generates about 1 gigabyte of data every second. That’s the equivalent of sending 200,000 plain text e-mails or uploading 100 hi-resolution digital photos from a computer with every tick of a clock.
Granted, the vast majority of this data is used in real time to control or report on the functions of the vehicle and has no real long-term value. Receiving a thousand “Tire Pressure Normal!” messages from a sensor doesn’t do anyone much good from the broader perspective of data insights, so automakers don’t bother to store that data in the car or in the Cloud.
But some of the data is valuable, and if you multiply that fraction by the billion cars on the road today, it doesn’t take more than basic arithmetic to understand why Big Data is attracting so much attention in the automotive space.
“When you look at the numbers, it’s no surprise that Big Data is garnering a lot of interest from auto OEMs,” says Henry Bzeih, head of telematics and infotainment at Kia Motors America. “It’s also generating interest from companies that have nothing to do with production-related OEM situations – companies who work in the automotive aftermarket, companies who offer solutions to the surrounding ecosystem.”
And yet, despite the volume of data available and the variety of interested parties (not to mention all the buzz in the media), there’s been far more talk about automotive Big Data and its potential than there has been action. “Comparing the amount of discussions around Big Data and the opportunity enabled by it, the actual degree of monetization is still extremely low,” says Robert Kempf, vice president at Symphony Teleca.
“Most people don’t understand what makes data big,” adds Scott McCormick, president of the Connected Vehicle Trade Association and industry adviser to the U.S. Secretary of Transportation. “A car may produce an exabyte of data a year (a billion gigabytes), but most is completely meaningless. Isolating the megabyte of data a month that’s really valuable, and then figuring out what you can do with it, that’s the challenge of Big Data.”
What carmakers can do with Big Data
Let’s leave aside the technical hurdles involved in analyzing Big Data for now (we’ll detail this in part II of this series), and start by considering what Big Data enables from a carmaker’s perspective.
On the simplest level, processing Big Data effectively allows OEMs to identify and respond to system-wide problems in a faster and more-cost effective manner. To illustrate, let’s take a standard 2013 model vehicle with an infotainment system in the dashboard. And let’s imagine it features a weather app that displays the latest forecast. Only today, when our driver gets in her car, she discovers that the app is frozen. All she can see is sunshine from the weekend, even though it’s Monday, and rain clouds outside her windshield are spelling rain.
In a non-Big-Data world, if the problem persists, our driver has to head into the dealership to get the weather app fixed, as does every other driver who’s experiencing the same problem. If the problem were serious enough (imagine the infotainment glitch causes engine failure), it could lead to a recall, which would represent a significant cost burden for the OEM and inconvenience and frustration for the customer.
Big Data won’t necessarily do anything for our driver on that overcast Monday morning. What it will do is allow the OEM to see if this problem is occurring in a certain region or is common across all regions. Furthermore, it will allow the OEM to detect if a specific sequence of activities or patterns of failures is triggering the glitch, without recalling all of the problematic cars or waiting for them to come into the dealership. In this case, perhaps our driver changed a channel on her Internet radio app, then went to the weather app, and the interaction of the two apps somehow caused the system to freeze.
Big Data and predictive analytics “enable a faster, preventive action,” Kempf says, “and might lead to being able to fix issues by a simple over-the-air software update, a small repair during the standard vehicle maintenance dealer visit or at least a recall with a limited amount of vehicles that are highlighted by the system as particularly exposed due to their usage profiles.”
“This is something we did not have before,” Bzeih says. “This is different, this is real-time. It provides much more information and much more concentrated information to improve quality.”
It’s not just customer satisfaction that OEMs are after. It’s a better bottom line thanks to reduced warranty costs.
Big Data opens up a host of other possibilities, many of them more proactive than the wait-until-trouble-strikes-and-we’ll-quickly-mitigate-the-problem approach.
For example, a second use case is to leverage Big Data to understand how customers actually use their products. In most cars, some features go untouched while others are highly valued. “Through Big Data, the OEM has the chance to learn much more about how customers use the vehicle and what their preferences are,” Kempf says. “Through Big Data analytics, the collected data can be transferred into a recommendation for roadmap feature planning.”
“That a subset of the customers take multiple short trips in a day, or long drives weekly, or have frequent treacherous winter drives, is useful to know so that [automakers] can advance their offerings to more fully satisfy those customers [or] potential customers,” McCormick says.
In addition, OEMs can start to segment their customers to see what value-added services might appeal to specific demographics within their customer base. This sort of personalization has already pervaded online shopping sites like Amazon.com – those who use the service regularly receive tailored recommendations based on what they’ve looked at and purchased in the past. Harnessing Big Data, auto OEMs can do the same.
Sales reps driving alone across large territories may not be interested in special discounted pricing on movies that they can’t watch on their backseat screens. A mom of four, on the other hand, who would love to distract her children with the occasional entertainment on the way to the grocery store, might be far more interested. “In many areas, the Big Data approach can help find the right pricing and the right target,” Kempf says.
OEMs can also start to share data with outside organizations to generate new revenue streams. The most obvious example here is partnerships with insurers who would like to collect a specific subset of data from cars – such as speed, time of driving, harshness of braking and cornering – to determine how drivers actually drive, rather than to base their risk calculation on far less reliable sources, like their age or credit status.
Insurers in the U.S. and Europe have already come to market with aftermarket solutions that enable the collection of this data – think Progressive’s Snapshot in the U.S. or insurethebox in the U.K. – but OEMs could also provide it directly to insurers, creating an appealing service for their customers as well as a revenue-sharing opportunity.
“More and more OEMs offer insurance services through partners and are thus interested in data that can provide them sufficient detail to calculate risks,” Kempf says.
(For more on insurance telematics, see Industry insight: Insurance telematics.)
This leads to a final use case in which OEMs can sell their Big Data to interested third parties. The U.S. Postal Service has put a price tagon its Big Data, for instance, enabling organizations to access the National Change of Address database for the sum of $175,000 a year. For car manufacturers, this may open up more privacy concerns than they wish to deal with, but whether it be insurers or advertisers or others, there’s little doubt that car Big Data could attract interested parties.
“Does the automaker need to use and sell all of its information?” McCormick asks. “No, but they do need to understand what it is they have. They might be able to sell it. They might be able to use it for their own purposes. They might be able to understand diagnostics, safety and quality to a degree they couldn’t before. Big Data opens all of these avenues.”
Part II of the series will examine why automakers have done little with the Big Data opportunity to date and the challenges they must overcome moving forward.
Andrew Tolve is a regular contributor to TU.
For all the latest telematics trends, check out Telematics Japan/China 2013 on Oct. 8-10 in Tokyo, Telematics Munich 2013 on Nov. 11-12 in Munich, Germany, Telematics for Fleet Management USA 2013 on Nov. 20-21 in Atlanta, Georgia, Content and Apps for Automotive USA 2013 on Dec. 11-12 in San Francisco, Consumer Telematics Show 2014 on Jan. 6, 2014, in Las Vegas.
For exclusive telematics business analysis and insight, check out TU’s reports: Telematics Connectivity Strategies Report 2013, The Automotive HMI Report 2013, Insurance Telematics Report 2013 and Fleet & Asset Management Report 2012.