Neil Tambe

I'm a Detroiter who happens to enjoy writing, national parks, orange juice, the performing arts, and fanciful socks. More than anything though, I aspire to be a good husband, father, and citizen.

Future of Auto #2 - Imagining Data (not tailpipe) Exhaust

I want to intern in Detroit this summer and I'm extremely interested in things like Consumer Insights, New Product Development, Future Trends Analysis, and Strategic Planning. Basically, I like building and launching new things. Seeing as how working in Auto is the likeliest of these routes (though not exclusively), I figured I'd see if I could actually come up with visionary ideas about the automotive industry.

This post is the second installment in a series I hope to keep with over the next few months. In it I will try to empathize with different customer segments and think of new products or services that would serve them in fresh ways. If you think my ideas are legit, I'd appreciate your help in finding a sweet gig for the summer. If you think my ideas are far from legit, I'd appreciate your feedback.

Note: I'm getting really excited about Auto. It seems like there are lots of ways to innovate this industry. It's wide open. The key seems to be to focus everything other than the car itself.


Update: As it happens, some friends who are more intimately involved with auto let me know that the transmitter portion of this idea already exists, but that the analytics is still underdeveloped. Though it makes this post a little less valuable, it probably makes the potential for a business a little bit more possible. Maybe one could collect more sophisticated data if necessary but then emphasize the analytics component.

My friend Cameron put it well, "The ability to capture data has far outpaced the ability to do useful stuff with the data."


Original Post: Tonight, I'm trying to get inside the head of people who manage fleets. Those people don't drive cars, their drivers drive cars. But, can an auto company provide them added value? I think so.

As I understand it, fleet managers probably want to know two things, mainly: how much their vehicles are being utilized and whether the cars need maintenance. Given this insight, it seems like there's lots of data that could be captured to help address those needs, with little inconvenience to the fleet manager or the driver.

Imagine this: an OEM installs a transmitter on every one of their fleet vehicles. This would tap into data about:

  • The vehicle's condition: e.g., it's oil level, fuel level, engine wear, or any data that is already produced by the car or could be cheaply captured with an add-on sensor
  • The vehicle's location: i.e., with GPS
  • The vehicle's use: e.g., RPMs, acceleration, g-forces of turns
  • Anyone interacting with the car: e.g., a driver driving the car, a mechanic
The OEM then transmits the data wirelessly to the cloud and runs analysis on individual cars and the health of the fleet. It could then sell this data, via a subscription service, to fleet managers looking for better information to manage their fleet of vehicles. For sophisticated client needs, maybe the OEM has a consulting business to develop customized metrics.

Maybe you could extend the use of the data by predicting when the fleet manager will need new parts for maintenance (e.g., when a vehicle needs an oil change), or by optimizing routes, or recommending ways for drivers to operate their vehicles more safely and efficiently.

Moreover, the OEM could open up the API to the transmitter's hardware allowing aftermarket sensors to transmit via the pre-installed transmitter. That way, the OEM's client could add data to the client's subscription, making the service more valuable for the client, which in turn boosts the OEM's positioning. Maybe you could even design the data in such a way so that it integrates into inventory or logistics systems already used by the client.

So now, we have a decently feasible business which has low capital costs (compared to producing a car) and the potential for ongoing returns. But wait, it gets better.

If you're an OEM, you now have lots of data about vehicles and their operation in the real world. You could mine that data to improve the production of automobiles themselves. Maybe you can detect defects or performance issues. Maybe you could see usage patterns and identify a new, underserved, customer segment. Maybe you could even scrub the data so it could be used by researchers to improve society.

But wait, there's more. If you perfect the technology on fleet vehicles (or even cars that you've leased to individual consumers) maybe you can develop a consumer metrics business which helps people improve the use of their automobiles in their own lives. Maybe you can recommend carpools. Maybe you help consumers have conversations with their kids who are new drivers. Maybe you can help people know when the most efficient time to go to the grocery store is. The possibilities seem limited only by the data that's collected and the imagination of the analyst.

This is fun.