IoT GUIDE

In an industry where technology and user preferences are quickly changing, tier-one vendors can steer their strategies toward higher profits by looking at how car data can drive new applications that better serve drivers. As we all know, cars today use data — either historical data captured in data centers or real-time data captured as generated in-vehicle — but some of that data has much more potential value when it comes to what buyers will pay for.

Consumers are used to data-fueled Amazon-style preferences but what’s coming down the road to cars is this concept married with real-time data that will open up a new universe of exciting solutions for the new world of the software defined car. Increasingly, drivers want to re-create the always-connected environment of their homes within their cars and smart tier-ones will build new apps on the data that enables such products.

This minute focus on data as the foundation of profitable future applications isn’t yet common but some of us in the industry are thinking this way now. Diving into all the ramifications of data and what it means to the auto industry is what innovation-focused Over-the-Air (OTA) platform vendors like us at Movimento think about every day.  We collect data, identify what’s mission critical, assess the transactional value of data and so on, thus we’re consumed with turning data into usable, profitable outcomes. So should you. Here are some simple rules to help guide you.

Rule One: Time-sensitive data is worth more

In many industries, real-time, time-sensitive data has more value in monetization terms than historical data sitting on a server. How so?  Let’s look at the most potentially lucrative kind of data — real-time data put to use in a single car. Real-time accurate traffic information and route guidance such as that provided by Waze has much higher monetary value to a consumer than traditional GPS-assisted route guidance.

This is the realm of tomorrow’s truly connected software defined car and the self-driving cars that are now in development by companies like Google – who has already acquired Waze.  Although it will be viewed in the future as somewhat primitive, today’s adaptive driver assistant features such as adaptive cruise control, self parking and similar elements are the baby steps of future solutions that will exploit real-time data for individual cars.

Rule Two: Multi-vehicle apps are valuable, but not quite as much as single-car apps

Time-bounded data in the context of many vehicles versus just one also has monetizing value, but not as much as apps aimed at individual cars. At a larger scale, real-time data that pinpoints how multiple drivers are behaving in a certain geography or other grouping would be useful information such as traffic information or accidents but future applications exploiting such data won’t bring in quite as much revenue as single-driver apps. Those rudimentary traffic-tracking apps that are available now don’t currently bring in high profits. As these apps get better, their monetized value will increase.

Rule Three: Behavior learning has higher potential monetization

Smart technology is already being used in cars, but we’re just at the beginning, think of the self-driving cars now being tested.   It’s useful to differentiate these technologies between machine learning (learning about the car) vs. adaptive learning (learning about the person). Think of this as learning that relies on historical data that takes longer to achieve as opposed to on-the-fly immediate information — also called behavior learning or adaptive learning — that quickly grasps individual behavior like how a person drives and uses that data to deliver personalized benefits and support. Imagine a car that suggests calling your doctor’s office, if your calendar is synced with the car and you are in transit to the location but are going to arrive late. Or a car system that suggests indoor restaurants for lunch because it knows you don’t like sitting in the sun.

Machine learning at large scale has the lowest lucrative value and involves things like amassing driver statistics and compiling car and parts statistics for market such as insurance or predictive maintenance. However, adaptive learning about the person has more potential value, particularly for apps aimed at individual drivers or cars.  Such behavioral learning can power apps that deliver immediate accident reports delivered wirelessly, or warnings of imminent security attacks, or alerts about driver behavior that could lead to collisions. Think about it — a driver is getting drowsy but there’s an app that involves behavior learning while utilizing real-time data. Getting such information immediately has higher value to drivers than less-specific, less-immediate data.

Rule Four: Historical-data-driven apps for individual cars have higher lucrative monetization value than those for many cars

Although real-time, single-car apps utilizing behavior learning are the monetized sweet spot, this doesn’t mean that historical-data-driven apps don’t have earnings potential, it’s the key to huge saving potential for OEM, tier-1 suppliers and also for consumers. Imagine the cost savings for OEM and tier-1 as a result of diagnostic, prognostic, preventive to predictive analytics of vehicle data before a cybersecurity breach like Chrysler 1.4 M vehicle recall or VW software scandal.

Consider the usage-based insurance market, which exists today and is making profits for companies like Progressive and Allstate.  They give customers an in-car dongle to capture driver behavior that can impact driver discounts. There are many more potential apps that exploit historical data undoubtedly coming down the road, since such software isn’t as complex to design as those requiring real-time data.

Insurance companies are also actively participating in the lower-monetization arena of large-scale apps using historical data. This is the area with the most here-and-now solutions available, largely from companies using historical data to understand how to price their products based on behavior, or certain zip codes, or driver age and gender.

Rule Five: It’s all about the data!

This entire framework is driven by data, which will be the foundation of all the lucrative apps of tomorrow. Although I’ve listed this as the fifth rule, it is implicit in all the rules and demands a broader view toward data and its sources than exists today in the auto industry. Tier-ones who want to position themselves for the future should now be lining up key partners who enable data collection — within the software defined car, in the cloud or both — and offer APIs today upon which profitable apps and services can be built.

These data providers are every bit as important as the advanced hardware and software in automobiles that use it.  Today’s car has the computing power of 20 PCs, more than 100 million lines of code and processes up to 25GB of data per hour. But without rich data, automotive machines won’t become truly connected cars. Besides future autos with much more computing horsepower than today, there will be a new data marketplace that will enable all the profitable apps that will boost tomorrow’s bottom line and work seamlessly with the new software defined world.

This article was originally published on LinkedIn.