The Internet of Automotive Things – Part 3 : Business to Business
In my last post we considered the impact of driver-centric connections orchestrated by the Vehicle. This time we’ll look at the car connected across it’s on board systems, and how this level of connectivity extends to the business relationships among the material and software providers. There are a lot of moving parts, and I’ll confess in advance it’s the area I find most interesting…
There are many seeking a seat in the car, some traditional industry players and many new entrants attracted by the smell of profits, not exhaust fumes. How these parties collaborate will dramatically affect adoption, attractiveness, and value realized across the ecosystem.
There’s no dispute that the OEM’s will continue to drive. Essentially serving as the “channel master”, it’s safe to say that OEM’s will add systems integrator to their portfolio as the information technology content of a vehicle increases. And it will. I’ve seen quotes estimating current software content at anything from 20-40%, and that a fully autonomous vehicle will be 90%. No one really says how that’s measured – Cost? Engineering time? Warranty claims? Sure not weight! But there’s no disputing that there’s a lot of code on board, and it’ll only increase.
It is, and will remain, the OEM’s job to make sure all that on board stuff plays nice. It is, and will remain, a challenge. I’m sure most recall the old – and Snopes-debunked – GM press release in response to a Microsoft claim on the relative advantages in the two industries. Kidding aside, electronic systems already represent one of the most significant warranty and serviceability issues and this will only increase as vehicle systems morph from mechanical to mechatronic. Haven’t seen much on “drive by wire” lately, have we. And as one who has just been presented with an $800 bill for a new Engine Control Module, I’m thinking that it had better save a lot of fuel relative to a carburetor to amortize the cost. Reliability, serviceability, and cost will be as important as convenience and safety in driving broad adoption of connected vehicle applications.
As we move into the future, this will impact not only the currently connected systems, but many more aspects of the vehicle. Some will be necessitated by increasing autonomous driving or assisted driving features, some by driver comfort and convenience, and some by safety and maintainability. Divergent standards could create havoc. Fortunately, SAE (Society of Automotive Engineers) has long experience in promulgating collaborative standards for vehicle technology, and they’re going be no less important in connectivity.
But there’s another aspect to vehicle systems that I haven’t seen discussed as frequently – how much intelligence over and above what’s necessary to enable operation needs to be on the vehicle. The amount of data collected in real time as a vehicle operates begs the question of how much, and how often, this data is shared with an entity or entities the vehicle is connected to. And can the frequency and content be dynamically adapted by any intelligent node?
We have today use cases where comprehensive real time telemetry is sent from race cars to engineering teams, and a constant stream of data is needed here. But for an ordinary passenger car, the burden on bandwidth suggests something less. The tradeoff is that the ability of the OEM and other partners to use big data technologies to assess vehicle metrics is predicated on sufficient data to tell a story. For comment’s sake, let me hypothesize that in addition to some cadence of polling, we will see some level of on board computing on the vehicle itself, perhaps with a subset of “big data” technologies to capture and keep history for some period of time, monitor behavior on board, and determine when a situation merits sending data back to the mothership. The two systems acting in collaboration could then determine whether to ask for additional history or more frequent sharing going forward. Seems reasonable to me, what do you think?
Finally, we’ll need to consider sharing of data collected from a vehicle. This will put the OEM in the position of being an “information broker”, as suppliers of vehicle components and systems who are asked to include sensory or intelligent devices in their products will expect to see not only the data from their content, but likely also collateral information to assist in problem resolution, warranty, and quality improvement processes. This seems rational, but I’ll bet will lead to some interesting discussions in the procurement offices of the major OEMs.
Gosh, this is going to be fun!