The IoT Trap

I’m sure that you’ve heard about the Sonos speaker debacle. (If not, read about it on Hackaday.) Basically, a company that sells a premium Internet-connected speaker wanted to retire an older product line, and offered a 30% discount to people who would “trade in” their old speakers for new ones. The catch: they weren’t really trading them in, but instead flashing a “self-destruct” firmware and then taking it to the recycling.

Naturally, Sonos’ most loyal customers weren’t happy about intentionally bricking their faithful devices, a hubbub ensued, and eventually the CEO ended up reversing course and eating crow. Hackaday’s own Gerrit Coetzee wrote up our coverage and mentioned that maybe Sonos just couldn’t afford to support the service for the old products any more, and didn’t want them to remain in the wild. So much so, that it’s worth 30% of the cost of their current product to get out from under the implicit contract.

By buying one of these IoT devices, you’re paying more money up front for the promise that the company will keep supporting the service that it relies on into the future. But providing this service costs money, and as more and more “products” are actually services in disguise, we’ve seen case after case of working machines shut down because the company doesn’t want to keep paying for the service. It doesn’t seem to matter if the company is small, like Sonos, or an immensely wealthy monopoly player like Google. Somehow, the people planning these products have a much shorter lifetime in mind than their customers do, and fail to make the up-front price cover costs.

This puts these companies in a tough spot. The more a customer loves the device, the longer they’ll want to keep it running, and the worse the blowback will be when the firm eventually has to try to weasel its way out of a “lifetime” contract. And they are alienating exactly their most loyal customers — those who want to keep their widget running longer than might even be reasonable. Given that this whole business model is new, it’s not surprising that some firms will get it wrong. What’s surprising to me is how many fall into the IoT trap.

So take this as a cautionary tale as a consumer. And if you’re in a company offering a product that depends on a service to continue to function, ask yourself if you’re really going to be able to support it for the customer’s idea of the lifetime of the product. What looks like a great deal at a five-year horizon might bankrupt your company at ten. Will you, or your customers, be willing to throw their devices away? Should they be?

This article is part of the newsletter, delivered every seven days for each of the last 210 weeks or so. It also includes our favorite articles from the last seven days that you can see on the web version of the newsletter.

Want this type of article to hit your inbox every Friday morning? You should sign up!

65 thoughts on “The IoT Trap

  1. A simple solution to a lot of the problem is to ensure that the end user can still have the device be usable locally.
    Ie, give the end user the option to run their own server for their IoT devices to talk to instead.

    For devices like printers, speakers, cameras, and most home automation, having a local server is a rather logical conclusion to be fair, and usually how things are running in the end regardless….

    Though, the trap is more about software updates, and cloud related services, something that isn’t always logical to begin with.

    There frankly is little to no reason for most devices to be connected to the internet. Pushing software updates is “nice”, and fixing security flaws is something one should expect a serious manufacturer to do. But at some point, the whole idea of “IoT” falls apart fairly spectacularly…

    The advice: “And if you’re in a company offering a product that depends on a service to continue to function, ask yourself if you’re really going to be able to support it for the customer’s idea of the lifetime of the product.”

    Can also be followed up with the question, “Does your product really need that cloud service?”

    1. Software updates for security flaws are necessary only for things that are connected to the internet, if the things were not connected to the internet in the first place, you may not need to fix potential security holes, therefore obviating the need for service from companies that do not want to give it.

      Of course, if one were able to run its own services, as you propose, that would actually solve something. Moreover if companies were forced to open source their designs when they go bankrupt or decide to stop servicing their products, that would be even better.

      Given how much we know about pollution and damage to the environment caused by many of these technological gadgets going into deliberate and planned obsolescence, it would be logical to force the manufacturers of any good to either recycle or to allow other people to do the support (open sourcing), just as manufacturers of bottled water, yogurt or other stuff in Germany are supposed to take back the packages for recycling.

      1. Requiring a manufacturer to out right open source their product at the end of its life is honestly a bit harsh from a business perspective. Since it can greatly impact a company’s future ventures, due to competition.

        (There is companies that after all don’t really do any development of their own and simply copies what other companies are doing, competing rather heavily with the companies that actually invest in the development due to having a lower price on their product, and risking that the original developer goes out of business. Enforcing this behavior is likely not beneficial. (Though not saying that companies investing in development always do the right thing, they too can be jerks…))

        Providing repair documentation, and software for one to use locally, and preferably documenting how to communicate with the device (API), is though still a rather logical thing to do.

        A product shouldn’t end its life due to a company not wanting to support it any more, but rather, the product should end its life when it frankly isn’t usable/repairable any more.

        That a company doesn’t want to support certain web connected features, and ends support for those features is acceptable. But intentionally bricking the product so that it can’t be used at all, is just rude…..

        For IoT devices, it wouldn’t be unreasonable to requiring that the manufacturer provides sufficient documentation for an end user to use the products “main”/basic features for as long as the product can be repaired, wouldn’t be unreasonable.

        But yes, a lot of devices don’t really have a need to even be “supported” by the manufacturer to start with.
        Running local servers, and interfacing with API’s makes a lot of sense for a lot of applications.

      2. Kind of disagree here a little. While the risk of a device’s security is directly related to being exposed, and unfiltered, to the internet; make no mistake, vulnerable devices can… 1) be used to move laterally, 2) Keep an uninvited guest present on your network while you are still trying to figure out how they keep getting in, 3) Try getting an embedded attacker out of a no longer supported device; it’s going to require some luck or serious hacking skills, 4) Depending on their place in said “network” be used to observe internal network traffic that, by your definition, is assumed to be less at risk. It is not to say such a device cannot be made relatively safe using compensating controls, but ask yourself realistically, how may people understand how to supplement a device’s security when they are likely facing multiple future flaws whose natures are as yet…. unknown. Worse yet, how many such devices have configurable firewalls and the ability to set IP based ACLs as part of a compensating regime to give the device, at least, a fighting chance?

        I do agree that they should make the device programmable to point to another server (or servers), offer the server software in a scaled down form and open source the server and firmware, best though, when someone (or a group of someones) agrees to do basic maintenance on both. Or just have a local use mode that is enabled as part of the decommissioning. (And should also include configurable firewalls and IP-ACLs; and yes, neither of these is perfect, but, better with them then without them, IMO)

        Lastly, the original warranty agreement should include a section regarding the use of alternative servers/local mode when EOL, in that, once enabled, your warranty is voided and no longer subject to support. This should clear them of any contractual obligations regarding support; if this, as the article suggests, is one of their worries (aside from the obvious of ensuring they keep getting those sweet sweet repeat greenbacks from hostage customers)

        Lets face it… most of us here void warranties…. and we are ok with that and expect it as an outcome of messing around with things. Who cares about voiding the warranty, really??? Besides, show of hands, how many people here don’t open a “thing” up, sometimes, until after the warranty is already expired???

        When it comes right down to it, I agree with Elliot, the costs of everything mentioned above are not likely to put a dent in their quest for more cash and will earn them some hard to come by good will from the exact same people that are likely to convince other people to buy the product. Even so the die-hards and hacker types are probably a small population among their customers. Throngs of their customers are not likely to abandon buying a new one vs taking the time to hack at the old one. Not everyone is a hardware hacker, nor interested in complicated doo-hickies they already barely understand.

  2. ” But providing this service costs money, and as more and more “products” are actually services in disguise, we’ve seen case after case of working machines shut down because the company doesn’t want to keep paying for the service. ”

    And yet some games get around this idea by having servers customers can run on their own long after the company has nothing to do with the game.

    1. And that is a good solution that sadly isn’t common enough.
      It is a great idea though – While the device/game/etc is new you get the superb performance (hopefully) of a well supported global/regional server with professional levels of uptime – which should keep company and customer happy. Once its obsolete but you still want to use it being able to tap into another provider or DIY home solutions are a nice way to keep the customer from being hugely pissed off while avoiding the continued costs of said professional servers – so everyone is at least not actively unhappy..

      Still for my money avoid any online/subscription service unless you can still get enough functionality without it

  3. These kind of implicit (or explicit) long term promises are what many companies end up trying to weasle out of. Take GM for example in the USA. They found that taking care of pensions for retired people was becoming more of a burden than they wanted to deal with. Personally, I expect any IoT device that depends on a “cloud” service to be a high risk item and to become useless, either sooner or later.

    1. It was more than GM could afford. Their retired workforce had reached nearly a million, and they only had 100,000 employees. Those 100,000 had to produce enough profit to maintain the company *and* keep the pension fund afloat.

  4. Long ago I bought a squeezebox, and returned it back to the store because it was not possible to play my own music on the thing without creating some kind of useless internet account.
    I have no pity for the companies who do stuff like that or the consumers who fall into that trap.

    About squeezebox…
    I heard some rumours that later the protocol was opened and other servers for that service popped up, but by then I had already lost interest in the thing.

    1. Long ago we got a Logitech Squeezebox and Logitech still updates and supports the free server software (at rebranded as “Logitech Media Server” which I run in a Docker and allows me to play my own music. There are free desktop and phone clients for it so I can play the music from my Windows pc. There is even an Alexa skillset for it. Synology has a Logitech Media Server package. I don’t know why Logitech continues to support the hardware, long after they bought and deep-sixed Squeezebox but I am grateful not all stories end up like Sonos.

      There was a firmware release shortly after we got ours that locked the box to their cloud service, but the next release opened it up again. You can probably find downgrade instructions from that time, people were buying the boxes and installing older firmware to unlock them. Sadly you could have held onto yours and it would have gotten better on its own but I think you made the right call at the time — how often does a company open up a product?

      1. I bet its still supported because some of the people working there want to still use theirs…
        As for company’s opening up the products that is damn rare. But if it is a good/popular enough device the community usually finds a way to keep it going..

  5. Every bricked Sonos speaker was bricked at the owners direction to secure a discount.

    Pick your villain: Sonos for offering the discount, or owners for choosing the discount, but honestly, the only people upset are the ones that can’t profit off reselling the old Sonos speakers.

    The moment a Sonos owner can get more than 30% of the price of the new speaker for their old speaker, they’d choose that option to the Sonos self-destruct discount.

    1. That’s true. The point I was trying to make isn’t about good vs evil, but rather how Sonos finds itself in this position where it’s willing to offer a big discount to get out from supporting its old hardware.

      In some sense, it’s no different from Microsoft, who’s telling you that they’re unwilling to support Windows 7 any more, and were offering a free/cheap upgrade path to their new software to get out of the burden. Only there, you don’t have to recycle hardware to do it, of course.

      So yeah. I hope this didn’t come off as Sonos bashing. I just wanted to point out that services masquerading as hardware are a strange deal, both from the consumer and the corporate perspective. I don’t think either really have it figured out yet.

      1. Microsoft too is trying to transition people to using software as a service. Their problem is that people buy the operating system once, install it, and then keep using it. As long as there’s third party drivers available, they never have to buy it again, so Microsoft -wants- to sell the service instead of the software to keep people paying indefinitely. The other alternative is to sell the operating system once, and then sell the updates and support, which would mean that nobody would buy the updates or support and you’d have unpatched computers everywhere.

        This is different from Sonos where the company wants to sell new hardware, so they’re putting the services in a loose noose and offering people discounts to buy new products that they’ve marked up +30% to begin with. They’re essentially blackmailing their customers into upgrading. Your choice: brick your device now and buy a new one, or it gets bricked later and then you have to pay the full price.

        1. That’s why after build 1607 Windows 10 only allows drivers that are signed with the newest method, introduced in Windows 8.0. You can still use the old drivers *if* they’re already installed in a previous version of Windows or in Windows 10 up to build 1607. So to do a clean install of Windows 10 using such drivers you have to first install build 1601, all the drivers, then upgrade to the latest build. That will keep the older drivers made for Vista, 7, or 8.x.

          If you try to install such older drivers in a Win 10 build newer than 1607, Windows 10 will allow the installer to go through the motions while silently blocking the files from being installed. If you try to do a manual install through Device Manager, Windows 10 *will lie to you* with a BS error message that it can’t find the specified file.

          So your best bet with an older computer (especially laptops) is to install build 1607 and all the drivers, then do an image backup so if some future build of 10 (up through 1909 is still safe) Microsoft decides to forcibly purge older drivers you can still go back to 1607 quickly then re-upgrade to the last build that allows pre-installed drivers with deprecated signing methods.

          1. Can’t you step into Advanced Options in Recovery and disable driver verification? That’s how I installed the software for a robot arm I had that had Vista/7 era drivers.

      2. You said:

        “In some sense, it’s no different from Microsoft, who’s telling you that they’re unwilling to support Windows 7 any more, and were offering a free/cheap upgrade path to their new software to get out of the burden. Only there, you don’t have to recycle hardware to do it, of course.”

        No, it isn’t anything like Windows 7 support.

        Microsoft had a very well-defined end-of-support date, and anyone that wanted support after that announced date was able to pay essentially private consulting rates to extend support their Windows 7 installations.

        Windows 10, and now Windows 11, are offered as free replacement OSS to anyone with compatible hardware and a retails/OEM software license (software assurance, or school/corporate site license holders are not included, as their current contracts allow for the use of the current software).

        MS did not try to end Win 7 support early, it ended right on time, as is their practice – they do not, and to my knowledge never said they would, support Win 7 ‘forever’.

        This practice is in-line with Linux (Ubuntu) and OS X (or MacOS or whatever they are calling it now), neither of which is supported beyond a certain End-of-support date. (Apple also drops hardware support, something MS never did until it imposed a TPM 2.0 requirement for basic Win 11 support, but quickly backed away from before GA date.)

    2. “Choose that option” assumes that buyers exist.

      Sonos simultaneously announced that “legacy” devices (as little as five years old) would lose unspecified functionality at an unspecified time, which made purchasing a used Sonos device an unspecified-risk proposition.

      I wasn’t paying attention to resale prices (and still ain’t) but I would expect resale value to drop – slowly at first, and then suddenly (after a cherished feature stops working, making this risk as visible as it deserves to be).

      1. Netflix and Hulu have gone through yet another round of changes that have disabled their service on older Smart TVs, Blu-Ray players and other devices. My little old 1080p Vizio from 2007 lost Hulu support in October 2018. Vudu still works but that and Flickr browsing are about all that still work on it. The games are dead because the backend support went away long time ago. Yahoo just doesn’t have real commitment to things. They bailed on their Smart TV software platform (which is what Vizio used in 2007) and after one initial burst of shows they axed Yahoo Screen – which was before anyone but Netflix was putting serious $ into original online video content. We’ll never know what was going to happen in Season 2 of “Otherspace”, directed by Kevin Feige and starring Karan Soni (Dopinder in the Deadpool movies) and Milana Vayntrub (Lili in many AT&T commercials). AFAIK Yahoo didn’t even try to sell off any shows to anyone else, they just shut it all down.

  6. “Given that this whole business model is new”

    It’s not though. The whole idea is to “add value” to a thing by attaching it to a service so you can ask more money for it – and then dump the service. Some would tell you to never assume malice before incompetence, but for so many of these companies the implicit point of tying the devices to services is to have the ability to brick it when they want to sell you a new one.

    The industry has a long history of planned obsolescence – anything that needs something replaced periodically, like a battery or a filter, etc. something simple that costs pennies, can be made such that only a special part made by the company will fit or work, and then you simply stop making those. Millions of devices bricked, and then the people have to buy the next model. Tying the product to a service is simply the same stuff in a new wrapper.

      1. Eletric cars have built-in obsolescence. A third of the price of the car is in the battery, which will die after 10 years. People who buy cars older than 10 years won’t pay more than a few thousand for the car, so they’re looking at paying twice the price for a used car to keep it running. You could fix it, but the car isn’t worth it anymore.

        1. Newer batteries are going to be longer lasting, and cheaper. So retrofits should be doable. Plus, there’s always somebody who only needs a 50 mile range for a vehicle so they can use it as is.

      2. Already enough of a fuzz replacing a control box on modern cars, now imagine replacing a battery charger box, motor controller or battery on a EV.
        You’re gonna get fleeced if the manufacturers and licensed service companies have anything to say.

        1. The law in many places is that you have to keep supplying spare parts for cars for at least 8-10 years after the warranty of the last model sold. This is part of the reason why car companies keep switching the lineup every few years – so no single model can build up a market presence to the point that people would come asking for spare parts 20 years later. If the car is rare enough, but not rare as a classic, people will just cube it instead because it’s too much of a bother.

          They learned their lessons from the old folksmobiles that had 30-40 year production runs and are still operating on the roads because they’re easy to repair and they were popular enough that third party spares became available.

    1. Quite some years ago I had an opportunity to pick up a bunch of network booting thin clients that used some proprietary OS. They were discontinued and the manufacturer would tell me *nothing* about them. They did have one page of generic information on their website but the list of available options with their part numbers was all in an image they’d deleted, and they got the web archive to delete from all their copies of the site. So I couldn’t even have part numbers or option names to search for. They would not give or sell me the old software to setup a boot server.

      They did offer to sell me their latest and also proprietary server and client system. Sorry no. “Piss off! Buy our new stuff!” is not a sales tactic that entices me to buy.

      So I declined the offer of the definitely useless thin clients and there went my idea to open a little internet cafe.

    2. Roxio GameCap HD Pro. It’s an HDMI and component video to USB capture device. The only way to get the software for it is if you already have the software for it, or at least whatever part of the packaging has the software serial number. It doesn’t matter that the hardware has its own serial number.

      If you don’t have the software serial number, Roxio’s stance is essentially “Screw you! You have to buy a whole new package!” just to get the software for exactly the same hardware you already own.

      Crap like that should be illegal. If a company wants to tie availability of software to hardware, proof of ownership of the *hardware* is all that should be required to obtain a fresh download of the software.

      There’s some open source software that’s supposed to work with this Roxio gizmo but I haven’t tried it yet.

      1. Laboratory instrument I support has a parallel port hardware dongle only supported under WinXP that prevents a software version newer than 4 versions older than the shipping version. We could purchase a USB dongle to run newer software, but the last version of software that supports our (now obsolete) instrument hardware is 3 versions old. And they continue to this day with encumbering their overpriced software ($2K a seat! for data acquisition and graphing/peak finding) with dongles. The instrument is $100K new. We are a hard drive crash away from losing this resource. It annoys the heck out of me….

  7. Not to be flinging conspiracy theories around or anything, but I do note with amusement that the current price of the Sonos Port, which is the replacement for what I have, the Connect, is almost exactly 30% higher than what I paid.

    1. Classic. 30% off a 30% higher sticker price! What a steal!

      What the hell do these IoT Sonos thingamajigs even do that bluetooth can’t do? Is it some dumb integrated AI assistant thing? Ugh, just get some normal bluetooth speakers people. Use your phone or something. Y’all KNOW that you won’t be using that AI assistant in a few months anyway, regardless of if it gets update-bricked or not. Nobody I have ever met still uses their AI assistants anymore. It’s so silly that we’re allowing these shenanigans for virtually no benefit at all other than a few moments of fleeting novelty.

  8. Whats odd is, Home Security Systems have been working for decades and not had this issue. Although, there is the monthly subscription fee. That’s really what’s missing from these systems is the thought that they can be maintained without a routine fee. But then, telling customers that is probably a show stopper for most people. Who wants to pay a subscription fee on something when a non-subscription version exists?

    1. Think again.
      People have also been left with useless or severely crippled “home secuity” systems after their subsription fee based company went out of buiziness, with no alternative to use the closed systems.

      1. Still better than simplysafe, an AI assistant masquerading as a security system lol. And one that can be defeated by a like $2 radio remote control you can get off ebay with no real technical skills involved. I have no idea why people buy those podcast products, they’re the worst. Also pretty sleazy how they love to advertise on all the true crime and serial killer podcasts. They’re so gross.

  9. It’s one of the reasons I go out of my way to not to buy any streaming videos, like from Amazon nor devices that require a service provided by some company. I just don’t trust them.

  10. This sort of scenario is happening with things like electricity meters (our old analog meters lasted 60+ years, will the new smart meters). Now companies can discontinue service (and offer an upgrade) as a continued cash flow scheme.

    What happens when this comes to Automobiles?

    1. One small problem with that example. You don’t own that meter, any more than you own the pole or the transformer sitting outside. Present IoT complaints are about what the public has purchased.

  11. Service-based or cloud-based products are specifically created as traps. Not only can they deliver a juicy monthly revenue stream from subscribers, but they also provide copy protection via enforced license scenes, and they can provide fabulous customer lock-in capabilities.

    Already have Sonos speakers? Switching to Amazon Echoes will doubles your costs, better to keep digging deeper into the Sonos pit. Already have Alexa speakers? Keep paying Prime for the music…

    For this reason, every IoT thing I buy I first try to find the open source, open protocol, and/or self hosted solutions. Some products have no good open options (garage door openers, thermostats, and laundry equipment all seem to be cloud based). Others, (such as sprinklers, home automation hubs, and televisions) seem to have open source systems, open source apps, or open interfaces on the devices themselves.

    I’m generally not trying to build my own hardware, or hack an ESP32 onto my garage door opener, or reflash my Dyson fan. I just want to be able to run my stuff offline, without depending on someone else’s cloud.

  12. There is a solution…

    When a manufacturer or programmer decides to stop supporting a product for any reason, it should automatically become open-source and community-supported. The later meaning that those using the product are then responsible for supporting it… or abandoning it as they so choose.

    The original owner would release to the public all documentation, designs and source code. They could agree to host that information for a period of time, but it would eventually either be taken on by some other entity or placed in some kind of public repository. Upon doing that, the former owner is absolved of all liabilities and responsibilities.

    Ideally, intellectual property laws would be changed to make this practice universal. But in the meantime, those who actually care about their users and their own reputations could make this common practice voluntarily.

    1. >it should automatically become open-source and community-supported

      One might be, the other isn’t guaranteed. Just look at “community support” on the various Linux distributions. Even supported devices often lose half their functionality because the people who can do the programming simply aren’t the people who are using these devices to any capacity, so they really don’t care if my printer can do borderless photographs or whether my sound system works correctly with all the speakers operating. “It prints, it plays, good enough!” In contrast, some very obscure cheap TV tuner card may be reverse-engineered to China and back because a bunch of geeks are using it for a cheap SDR. Of course if you try to use it to watch TV in Caffeine, you get a green screen and the kernel panics. True story.

    2. Exactly this! There’s a huge amount of hardware, phones in particular, that while perfectly functional are being binned because manufacturers stop giving them updates.
      Even if you don’t care about security updates, the developper model (for Apple for instance) is that you need to use the latest iOS, so clueless developers usually make their apps incompatible with older versions for no reason.

      Whatsapp for instance just disabled the app on my phone (a few weeks ago). They want me to update the system software, just that is not possible, so it stopped working. WTF!

      I would add that the HW design should be opened as well, that would allow people to keep making parts, or to be able to recover the parts. Recycling electronics currently amounts to crushing stuff up into powder, which is totally insane given many of these phones are functional, so all parts (the passives: resistors, capacitors and other generic parts, including flash memory) could be recovered and reused (note the difference with recycled!)

    3. Much better yet:
      Enforce open protocols only, and forbid proprietary protocols.
      This immediately stops the vendor lock-in and bings copetition to the companies who deliver those services.

      Problem is:
      I’s’ what consumers want. It is not what companies want.

  13. The “implicit contract” and dependence on a cloud built on an unsustainable economic model is one of the reasons I refuse to by products that don’t provide a local API and have a strong preference for products that run on Open Source code.

    Unfortunately, this is an environment where asymmetrical information dominates and therefore it is virtually impossible to get consumers to make informed decisions about these products. That allows these vendor-cloud-dependent devices to proliferate and worse, the perverse economic incentives of the surveillance economy actually provide a disincentive for companies to do the right thing.

  14. My perspective comes from home automation. There are now a flood of home automation products, switches, thermostats, curtain controllers, etc., that rely on cloud services. I don’t employ these. As noted in this thread, if the manufacturer of the product quits the business, your device doesn’t work any more.

    And then there’s a security risk as now, not only does Google/Amazon have access to your house, but also a third party. This more than doubles the security risk.

    And, as noted, you can be held hostage to subscription fees.

    So I’ll go with the open source server

    1. Would you rent a set of speakers if you could just own them?

      “X as a service” is a terrible prospect – it screws you over as a customer because it’s mainly used as a leverage to charge more for the same thing, and it doesn’t solve the fundamental problem where the company up and decides they don’t want to support the product anymore. It just makes it worse, because then you can’t even use it out of support – since it’s tied to some online account or server to have the excuse. Kinda like how Microsoft is trying to kill off local accounts, so you couldn’t even log on to your computer without having access to their servers.

      The price would have to be pretty very low because you have to look at the total cost of ownership. The company has to make the money in a reasonable time, so they would want to charge you $200 a month, and then keep on charging the same – but you’re looking at having the set of good hi-fi speakers for 15-20 years, so the best you’ll pay is about $20 a month and even that’s pretty steep.

  15. “Given that this whole business model is new, it’s not surprising that some firms will get it wrong.”

    Here’s what’s wrong: THE BUSINESS MODEL ITSELF.

    Back when the “Internet of Things” was being envisioned 30 years ago, we all pictured that there would be a modicum of local control over the devices you — supposedly — OWN. There’s no reason why a third party should control the connection between, say, my stereo receiver and my speakers and certainly not control it to the point where they can stop or interrupt that connection at will. Likewise, I don’t need “service providers” operating my devices in order to ration out cat kibble, open and close my garage doors, dim my light bulbs, speak to someone else in another room of my home, or adjust my thermostat or the temperature of the water in my shower.

    I should be able to buy a home automation system which I own and control end-to-end. I should be able to exercise granular control over who has access to it, what data it sends back to third parties who are openly identified to me. Any business model not based on transparency and openness makes me the product, not the customer. And until we all reject this one, we’re doomed to relive this Sonos debacle over and over.

  16. The best way forward, in my opinion, is to do what the Brits are proposing in their IoT legislation and mandate that any Internet connected device have a guaranteed lifespan during which updates will be provided by the manufacturer. If manufacturers were required to make this information public, some would logically take the next step and put the lifespan on their packaging and promotional materials. Over time, I believe customers would naturally gravitate towards the “10 year” doohickey over the “5 year” or “3 year” versions.

    In this situation, at least some companies would seriously contemplate continuously supporting their products or at least pushing out outlandishly long support times just for the marketing benefit. Companies that based their products on an open source model would see the strongest benefit.

  17. They knew they were intentionally crippling these products by tying them to a cloud service.

    It would be smart to design their service so that there is an easy way to disable the cloud service down the road as it reaches the end of its “service life” and then enable some fallback local option.

    Arlo cameras have an ability to attach a USB drive to the hub that currently allows a backup that you cannot access unless you remove the drive from the hub and attach to another computer. If they ever wanted to deprecate the cloud storage portion it wouldn’t be too hard to flash the firmware on the hub to turn it into a tiny NAS to enable any attached USB drive to replace the cloud storage.

  18. Don’t say things don’t need to be connected to the internet. It’s naive, antiquated and ignorant. And it’s always buy at your own risk. Routers are usually abandoned after a year or two. Sonos is a terrible example. They are a 1400 employee company who is largely irrelevant now. Of course they can’t maintain all the hardware they make. Common sense.

    1. Things don’t all need to be connected to the internet. Don’t tell me what to do. It’s not antiquated and ignorant, there are legitimately everyday objects that will never need to be networked, nor should they be. I detest that kind of forced progress attitude. Progress is not a straight line leading from now to one single future that is totally inevitable and only Luddites would desire otherwise. Progress is a three-dimensional space with a path that we can influence and choose. Change is certain, but the illusion that the change they’re hawking is the only inexorable way is not certain.

      I think people will get pretty sick of every appliance in their home having a mind-bogglingly cheap, badly-designed, fragile, and horrendously insecure networked functionality that provides very dubious benefits. Things will become more technologically advanced, but we’ve entered this weird postmodern assumption that technological advancement exclusively means internet and apps and phones and AI. There’s other things to choose and improve. The future doesn’t have to be a dumb consumerist hellscape where every single object in the world has a crummy wifi connection and has to talk to a remote server to do the most basic functions which have always been accomplished locally until now. It simply doesn’t need to be that way.

  19. Chumby and sony Dash come to mind. Chumby being somewhat an exception having its software available and accessible. Despite the company’s demise there is a cult of hackers who refuse to let Chumby die. Letting devices go onto the out moded pastures goes farther back than IOT. Intentionally sabotaging with software (or lack thereof) is not new. Adding the word ‘service’ is a misdirection. Internet does make the process easier and faster. Disabling ‘update without your knowledge or permission’ may be a violation of ‘service’ agreement.
    Forcing upgrades by failing to support is a business model that been around a very long time and not likely to dissipate.
    Thats why there’s hackers. Hackers decide when the value of an item is exhausted leading to villianizing by various concerns. Not to be confused with malicious activities of a few.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.