This week’s dive into big auto’s embrace of big data comes on the back of two recent news stories. The first is on the partnership between Ford and US insurance giant State Farm. The second is on the GM-backed start-up Wejo’s partnership with Microsoft, Palantir and Sompo. We will take a look at these news stories separately before discussing them together (in the personal comment) as part of the auto industry’s broader trend to monetise vehicle data.
Firstly, Ford and State Farm have announced they will be continuing a partnership to share data together now that they’ve successfully concluded a 1-year long pilot project. The companies claim that under the pilot certain owners of Ford vehicles with ADAS features (who were also insurance customers with State Farm) were able to save up to 20% on their insurance costs. Under the data sharing arrangement, Ford shares a vehicle’s fit-out of ADAS features with State Farm as well as its fleet data on how each of these features (or their combination) impacts the frequency and severity of auto claims. State Farm in turn is also sharing claims data with Ford to help inform them on how specific vehicle features in a variety of claims scenarios will impact the cost of those claims.
One of the data backbones that enables this arrangement is Ford’s Vehicle Build Data product which provides a comprehensive view of an individual vehicle's feature set, for which Ford has data for all its vehicles dating back to 2010.
“What's unique about this project is we're not just looking at the make and model of a vehicle, but ultimately to the individual safety features on each vehicle,” said State Farm Vice President of Operations, Craig Isaacs. The ability to access build data going back to 2010 also means that “second and third owners also benefit.”
Looking beyond just insurance we have the article about the GM-backed vehicle data startup called Wejo. At the end of June, Wejo announced partnerships with Microsoft, Palantir and Sompo Holdings to “improve its ability to collect, store and analyze data from millions of connected vehicles around the world.” Palantir is a developer of software and data analysis platforms, most notably (or controversially) perhaps for its contracts with the intelligence community in the United States to assist in counterterrorism as well as immigration operations. Put simply, the company builds programs that integrate massive data sets and spit out connections and patterns in user-friendly formats. Sompo Holdings is a global insurance company headquartered in Japan, with significant reach in the auto-insurance sector. And Microsoft of course is a technology behemoth that, for the purpose of this partnership, is providing its machine learning and AI capabilities as well as its cloud hosting platform for data collection.
The difference between what Wejo is doing with its partners compared to Ford and State Farm is that it’s focusing on connected vehicle data gathered in real time, rather than feature fit-out data gathered at the point of production. “We have 11 million live cars on our platform out of a supply base of about 50 million vehicles. We [have] engagements with over 17 OEM on the platform, and we’re processing 16 billion data points a day, a peak of about 414,000 per second…” explained Richard Barlow, founder and CEO of Wejo. Barlow says Wejo can see 7% of all vehicles moving around New York, 6% around California and 20% around Detroit from partnerships with automakers like GM, Daimler and Hyundai.
With this data, Barlow says that Wejo can either sell raw, anonymized data — collected from vehicles with the consent of the owner — to businesses, developers or governments, or it can perform data analytics for them. A third option also seems to be that Wejo and its partners can use the data to come up with their own products. Some initial application areas that Wejo have in mind were listed as traffic solutions, remote diagnostics, integrated payments, advertising, retail, logistics, as well as using the data to help companies developing autonomous driving technology.
“Wejo collects data that can recreate a car crash, from how each driver stomped on the brakes to which airbags were deployed to the speed of impact and which sensors were destroyed,” said Barlow, whilst indicating that this makes Wejo an attractive partner for companies in the insurance, auto repair and self-driving space.
Big data has been a buzzword in the tech world for a while now and its potential to play an important role in re-shaping the auto-insurance industry in particular has been touted for some time. Tesla was (and perhaps still is) the first car company to launch an insurance product targeted specifically at drivers of its own vehicles, with a focus on using real-time data customised to each individual driver as the foundation for determining risk and pricing. Earlier this year we also reported on moves from the broader auto-insurance sector into looking at integrating information about ADAS features into their own policy pricing frameworks – noting that “Testing of the data is already underway by motor insurance providers in the UK to understand how these vehicle centric insights correlate to claims.” Although that testing seemed to be happening with ADAS features on a more general level (i.e. common classification of ADAS) across the whole landscape of OEMs rather than between individual OEMs and insurance providers.
That seems to be the difference between what the article on Ford’s partnership with State Farm is touting here compared to what we covered earlier this year. Rather than an insurer treating an ADAS feature (e.g. advanced lane-keeping assist or forward collision warning) from multiple OEMs as having the same weighting in determining accident likelihood and severity, partnerships with individual OEMs can provide much more granular detail down to how that OEM’s feature paired with a particular car model from a certain year will perform. Taking this one step further, we have what Wejo is looking at with connected vehicle data from a fleet of 11 million vehicles (and counting) across several different participating OEMs – though even they don’t seem to be suggesting that they will craft insurance policies around the real-time driving habits of individual insurance customers (like Tesla is claiming to do) since Wejo’s data is still supposedly anonymised.
We have also been given a more close-to-market sneak peek at other ways beyond insurance that connected vehicle data may be monetised, with Wejo listing traffic solutions, remote diagnostics, integrated payments, advertising, retail, logistics and, of course, assisting companies developing self-driving capabilities. While the usefulness of connected vehicle data for managing traffic, logistics and developing self-driving cars is quite logical, what’s interesting to me are the less obvious of these applications – such as integrated payments. Could a company like Paypal for example try to leverage this data from Wejo to create a universal payment solution for all the different highway toll systems in the US? And how does Wejo plan to monetise this data for advertising and retail purposes? (and what would be the social acceptance and privacy implications of that?) As usual, announcements like these create more questions than answers. But usually when a sector moves beyond discussing potential uses of technology in broad ‘changing the world for the better’ startup speak to laying out actual strategies to monetise that technology, that’s a much clearer signal things are about to shake up fast.
Written by Bobby Chen
RISE Mobility & Systems (Elektromobilitet)