Ideas to keep data from the home, in the home

Last week, I asked a lot of questions about the network of surveillance we are building with the IoT, how much of it was coming to light in the wake of the Black Lives Matter protests, and how we seem to be building it out further to handle the pandemic. Apparently I’m not alone in my worries, as I’ve seen three responses to these questions emerge that might help address some of my worries: from a startup, a government regulatory body, and a group promoting the IoT in Europe.

Let’s begin with the startup. Kneron, a company I covered in April that makes a chip for edge-based machine learning, on Thursday launched a platform for sharing device data within a home network and performing local machine learning on those devices to provide services. Any device that’s part of the platform needs a Kneron chip, so I have my doubts about how far this can go, but the architecture is worth looking at.

The so-called Kneo platform lets any device that has the appropriate Kneron chip (although ideally, it could be any chip running the appropriate software over time) to become part of a secured, local mesh network where device data can be analyzed and acted upon. For example, a doorbell camera that has the Kneron chip inside could use that chip to run an image recognition algorithm locally on the camera. When the camera recognizes you, it could signal a door lock to open over the local network.

This becomes more interesting if the camera sees you with an unfamiliar face. It might then refuse to unlock the door, or would unlock the door but also send an alert to the network, depending on how you set it up. The network is secured using the blockchain, which is where the second aspect of this platform comes into play. The idea is to create a marketplace of apps that can use the Kneo data to perform specific tasks.

So the owner of the video camera and door lock might download an app that prevents anyone from following a recognized person into a building. Or someone who owns a smart light and a door lock might download an app that turns the lights on when the door is locked and the owner’s smartphone leaves the building.

Essentially what Kneron is proposing is a secure and local exchange of data that could govern a smart office, smart home, or even a smart city. The ability to share data with apps consumers choose presumably lets them control what data companies have access to, including what data they want to share in exchange for a discount or service. I’d argue that we sort of already have this ability today, when we click through end-user license agreements in order to get our hands on the latest data-scraping apps, but at least in this scenario, my data would stay on the device until I allow it to leave. And if I uninstall an app, I could exit the contract I had with the provider and protect my data again.

But I have also witnessed so many efforts to create these marketplaces already. Generally speaking, the hardware (or software) required should be widely used in the market, and either the app store should have use cases so compelling that they drive the purchase of the hardware or the pain point should be so great that people will flock to a marketplace just to escape the status quo. I’m not sure we’re there with any of those factors yet.

Meanwhile, the EU’s competition commissioner, Margrethe Vestager, has launched an antitrust inquiry into companies that collect and use smart home data. Rather than surveillance, the primary focus of the investigation is who can access data from smart home devices and how that collection of data might be used to advantage existing market players.

She was quoted in a news release announcing the inquiry as saying: “[A]ccess to large amounts of user data appears to be the key for success in this sector, so we have to make sure that market players are not using their control over such data to distort competition, or otherwise close off these markets for competitors. This sector inquiry will help us better understand the nature and likely effects of the possible competition problems in this sector.”

The release lists mobile devices, smart speakers, and wearables as targets of the inquiry and notes that it will focus on restrictions of data access and interoperability, as well as certain forms of self-preferencing and practices linked to the use of a proprietary standard. The EU investigation is less of a referendum on surveillance as it is an inquiry on lock-in and control of a smart home platform associated with user-generated data.

However, by opening up the lock that players such as Amazon, Google, and Apple have on the smart home, it also opens the door to alternative business models that don’t rely on surveillance. As an aside, the Project for Connected Home over IP standard proposed by the large smart home companies might solve this issue as well, although perhaps not in a way that reduces the level of surveillance.

Finally, we have a proposal from the IoT Council around user identity to create a framework that separates data from the individual creating it and linking it to that individual only when necessary. The proposal suggests that, much in the way Apple and Google proposed with their contact tracing efforts, when information is created it’s tied to a token and the token remains unaffiliated with the user until such a time that it’s needed for a specific purpose the individual has approved.

Rob van Kranenburg, founder of the IoT Council, suggests that the system of disposable identification is needed because the aggregation of data about individuals, their lives, and their health will result not just in omnipresent surveillance, but a world in which there is individualized pricing for every good and service. In other words, when I stop at a store for milk, my price would be set based on a variety of factors specific to me, including my ability to pay. Van Kranenberg worries about the loss of economic agency as much as the loss of freedom of thought and action, but I find his interest in the decentralization of identifiable data compelling.

These are three potential solutions to the problem of surveillance, especially surveillance as an inadvertent byproduct of a lack of competition, and there are likely many more. Keep them coming.