For anyone interested, there's been a response to this article by Ralf Rottmann: https://medium.com/@ralf/andreas-brekken-proves-he-has-no-cl...
What is the utility that IOTA provides, that Bosch are deploying at production, and where is this used?
You won't get a decent answer because as far as I can tell there isn't one.
What is the utility that IOTA provides?
There's nothing in your link that talks about the utility that IOTA provides. There's machine to machine payments, but that's not really utility because currently machines don't pay each other, users do. There's no actual demand for machines to pay each other. This isn't unsurprising because it's a new technology and sometimes technology has to develop the demand.
What are Bosch are deploying at production?
The link you supplied has nothing being deployed at production. There's talk of the XDK, but that's a developer kit, not a final product. If you go to the XDK site and search for IOTA, iota or tangle you get no results. This is not contingent with the importance that IOTA supporters place on the Bosch relationship.
Where is this used?
According to the link you've posted, there are a number of Proof of Concepts, but no actual deployments at production scale. It's not being used outside of PoC at this stage.
The thing is, you could've just tried answering the question but posting an AMA that doesn't answer it is frankly, shitty. If you believe it does answer the above 3 questions, at least do the decent thread and link to the comments with the answers.
>On the Internet of Things, edge nodes are mostly low-powered, occasionally connected devices, sometimes with just a few kilobytes of on-device memory. Nobody expects those devices to be in sync with anything. That is, why IOTA has a concept of nodes. Fullnodes and permanodes. These nodes stay in sync with the Tangle. Or, if you consider use cases which require offline-capabilities, even just with subparts of the Tangle, which can be attached back to the main Tangle at a later point...
Andreas seemed to be suggesting that no IoT device would be able to take part in the IOTA network, because his attempt to run a full-node required lots of data.
All the arguments about "how can something that's pre-alpha be worth billions of dollars" are perfectly valid, but the original article seemed to address IOTA's worth for IoT devices, and seems to have missed the plot on that.