UK IT Specialist Unable To Boil Water, Make Tea

In our latest episode of “IoT-Schadenfreude Theater” we bring you the story of [Mark], a British man who can’t boil water. Or more specifically, a man who can’t integrate MQTT with Amazon Echo, or IFTTT with HomeKit.

Yes, yes. We all love to laugh at a technology in its infancy. It’s like when robots fall down: it’s a cheap shot and things will surely get better, right? Indeed, the Guardian has had its fun with this particular WiFi kettle before — they’re British and nothing is more important than a remote-controlled cuppa.

Every time we hear about one walled-garden protocol not speaking to another, and the resulting configuration mayhem that ensues, we can’t help think that [Mike] was right: home automation has a software problem. But that’s putting the blame on the technology. (We’re sure that [Mark] could have made the kettle work if he’d just applied a little Wireshark.)

Strongbad's VCR
Strongbad’s VCR

There’s another mismatch here — one of expectations about the users. A water kettle is an object that should be usable by grandmothers, and a complex networked device is clearly aimed at techies and early adopters. Combining the two is asking for trouble. Non-functioning IoT devices are the blinking 12:00 of our generation.

What do you think? Where’s the blame here? Poor design, bad software stack, stupid users, or failure of mega-corps to integrate their systems together? More importantly, how could we make it better?

Headline image:Fredy Velásquez Orozco, via Wikimedia Commons Thumbnail image: Markus Schweiss, also Wikimedia Commons.

54 thoughts on “UK IT Specialist Unable To Boil Water, Make Tea

  1. After all his epic struggles he emerges victorious and while standing _within arms reach of the kettle_, calls out to “Alexa” to turn the kettle on for him. It’s a struggle every parent knows all too well. He _could_ just do it himself, but it is better for all involved to teach the youngster a life lesson…right?

    *mutters something about dying on hills….*

  2. Ever since reading Sparkfun’s excellent blog post on protecting the core functionality, I have started to see these IoT devices in a new light. If the little button on the front will get the kettle to boil water regardless of the state of my network, then fine – I have a reasonable backup for when things go south. But if IFTTT’s financial stability is a critical link in my ability to do something basic, mundane, critical, etc, then we have a problem.

    1. Agree totally!

      And I’ll go one further. If everyday devices are going to have the Internet inside them, they have to make the Internet part as accessible as the other functions too. This includes a big “turn off the IoT features” switch, a button to update/upgrade the firmware, and maybe another for pairing with devices/generating safe passwords.

    1. Sandisk tried this with their eye-fi SD cards by creating a software dependency on a web based server that was not functionally required. They planned to switch the server of so all the products would stop working. There was such a backlash that they released a new version of the software that had no server dependency and would run (perhaps) indefinitely. Consumers wonk put up with this sort of forced obsoletion.

      1. Well, until they start forming single product companies under various shells registered in teh Cayman islands with principals unknown etc etc, as inevitably will happen, because it can. Then poof, company all gone, no heads found to be rolled.

  3. This was a major reason behind my thermostat purchase recently. Ended up going with a RadioThermostat model because they have a well documented local API so I don’t need to depend on their cloud service continuing to work. I’d love to have been able to get an EcoBee or Nest as they’re prettier and a bit more featureful out of the box, but their walled garden approach is a non-starter (Just look at when EcoBee’s servers went down for the holidays). It’s a sad state of affairs when {Product Name} hack, {Product Name} root, and {Product Name} shell are my first google searches when shopping because I want to be able to use it.

      1. Exactly why real hackers program with butterflies. They open their hands and let the delicate wings flap once. The disturbances ripple outward, changing the flow of the eddy currents in the upper atmosphere. These cause momentary pockets of higher-pressure air to form, which act as lenses that deflect incoming cosmic rays, focusing them to strike the drive platter and flip the desired bit.

        If you use a text editor, or worse a /cloud based/ text editor, you can never be sure what’s really going on.

      1. Programmable is completely OK, if self contained. I prefer digital readout anyway, also on watches. Time is a number, not an angle of hands. But cloud based is a complete no go. Some small LOCAL home server as a core component, based on an xWRT Router or any FruitPi would be also fine. I bought some Sonoff devices recently, nly to find out, that they seem to need a cloud service out of the box :-( But they are open and documented, so that should is easily changed.

  4. “Non-functioning IoT devices are the blinking 12:00 of our generation.”

    I think that is an insult to how useful VCRs were before online options happened.

    Seriously they’re like someone woke up one day and said “I know what the world needs! A middleman between them and every on/off switch.” …. whoever that was, make sure they get their seat on Ark B.

    1. And yet so many people just could not use the timer or set the clock on a VCR, and never tried because the user interfaces in those devices – beyond the basic row of buttons – was absolute crock of shit and it reset itself back to blinking 12:00 with every tiny power failure.

      Besides, TV stations started acting wise and publishing program guide times so that anyone actually setting their VCR to record when the program was supposed to start would get several minutes of commercials on tape.

      A lot of the time you simply could not trust the VCR to start recording when it was supposed to, so you either had to stay up until the program starts, ask someone else to record it, or put a 4 hour tape in and record the entire evening.

  5. The problem here is greed, more specifically the desire of manufacturers to lock users into their own walled gardens. Instead of giving their devices some kind of open interface that anything can use, they use proprietary protocols that communicate with proprietary apps and virtually always also, for no good reason, a web service hosted by the manufacturer. This way users are forced to use the manufacturer’s ecosystem and it gives them an incentive to buy more products from the same manufacturer. The manufacturer wants their customers to buy all their IoT devices from one manufacturer, and they want to be that manufacturer. The result is this scorched-earth landscape of total incompatibility – and universally awful security, a closely related but different matter.

    Trying to mitigate the problem by bringing IFTTT into the mix is a laughable attempt at a solution which only compounds the problem with another proprietary web service which should have no reason to be involved.

  6. Engineers solved the blinking 12 problem with a standard to broadcast the current time and timezone offset so that as long as the VCR was tuned in the clock would be set automatically, but penny pinching appliance manufacturers ignored it because it would mean a fractionally more expensive chip in each VCR.

    1. The limitation was battery / cell technology. About the only thing that was rechargeable was lead acid (PB HCL) and they weren’t sealed. The brand GATES did make some but they were expensive. Unlike modern chips, old chips needed enough backup battery power to run the steam engine!

      1. Waaaaat? Nicads been around commercially since 1950s… though they may be referred to as D.E.A.C.s in older tech pubs. Brand name that got genericised a bit. Though NiCad is a trademark of SAFT also. Seen ’em in 70s gear. First RTC backups in PCs where non-rechargable lithium cells in early 80s, those last a lonnnng time usually. Or did.

        1. Now, when you say it, I remember a another rechargeable torch of my grandfather with “DEAC” cells in it. But in the 80ies the PC mainboards hat 3,6V NiCd backup batteries. The lithium primaries came in the 90ies

          1. No not CR2032s the lithium from before came in various sizes the small one looked about the size of a boxed up AA the largest like a 4 pack… Oh and some actually built into the clockchip for a few years. Look up Dallas clockchips for those.

        1. The alarm didn’t go off because the clock was made for a non-rechargeable battery which had long since turned into an acid smudge dripping out of the bottom of the battery case. And that is why VCR’s didn’t attempt to use battery technology of the era. Chemically stable rechargeable batteries were not a thing then.

          1. Bullcrap. Those same 9 Volt batteries lasted for years in smoke alarms without leaking. They were designed for 10 year shelf-life.

            I still have one clock radio dating to somewhere early 80’s and it keeps time when disconnected. It just blinks the display to save power.

          2. VCR’s came out in the 1970’s and back then you needed 100mA or more for a back up battery. Rechargeable batteries had a very low energy density and just couldn’t do that and neither could chemical batteries unless you don’t mind installing a truck battery into your VCR.

            The posts and pictures here are of 1980’s on onwards batteries and they wouldn’t have been able to supply the current requirements of old chips even if they *did* exist in the 1970’s. Go look at the Icc spec for the original 74xx series chips, VCR chips were even worse.

            It’s not exclusively about energy density, it’s more about energy consumption. This old tech didn’t have a standby mode. Then you switched it *off* it had no electrical supply at all. So if you watched one 2 hour movie a day then you had two hours of charge and 22 hours of discharge at a comparatively high discharge rate.

      2. Lead acid is Pb H2SO4.
        But anyway, there were sealed rechargeables at least in the 70ies. We had a rechargeable torch in the 70ies. Its battery was definitely sealed also. Probably NiCd, perhaps lead acid.
        And our first VHS recorder (early 80ies) had at least a 1hr backup. I had a portable CD Player which used a sealed lead acid battery in the late 80ies. It was funny as its cells looked like todays LiPo cells, wrapped in silvery foil.
        But it sure was a question of money, the VHS was one of the first with “HiFi Stereo” and it was really expensive.

  7. I see lots of people here complaining that IoT devices use proprietary/closed software and interfaces.
    But that’s what keeps them secure from hacking and DDOS attacks.
    Oh, wait…

    1. perhaps the same reason as the telephone was not adopted early on in Britain: I mean, we have messenger boys to run to the telegraph office, that’s good enough, who would want faster service than that. and have a bell to interrupt your train of thought at any moment!? preposterous!

      point is, you can’t always foresee the benefits until you try it out. and sometimes those benefits only appear after a sufficiently large scale of adoption.

      sure, most of the benefits today are for DDOS networks, more than the consumer….

      but even that is a lesson worth learning.

  8. Unfortunately this is the “enviroment” they would like to pack us tech people in. I think it boils down to control, if a whole town’s tech runs off a IOT Grid it would be simple to blackmail that town into paying more for the “cloud” service.

    I recently did the excersize where I wanted to have all our office server environmental data on a raspi gui to monitor temperature, AC status and backup load but realized Mqtt on a clobbered together local server is 95% more stable and easier to implement than a “simple” cloud service, it end up being like buying a new car……oh sir it does not come out with mirrors or electric windows……but we can add it to the package for simply $$$$$

  9. “Non-functioning IoT devices are the blinking 12:00 of our generation.” Made my day… Thanks :)
    Whatever happened to teakettles with whistles? Though I did find out the hard way that forgetting to attach the whistle can turn copper black, I still wouldn’t prefer an I[di]oT-interface for the vast-majority of my Things. Am OK with light-switches. Can’t reach ’em from the couch? Screw in some eyelets and run some string. Or try to teach the cat a trick or two…

  10. I just use IoT as a way to connect sensors to things more cheaply than before. I never bought any devices on the market, it’s easier to build them myself than research all the platforms out there and then go through setting them up (big problem with mass market appeal there?) then also dealing with interoperability.

    So for my use case, data from IoT devices flows towards me (rather than from me towards devices), and helps me make decisions about when to manually toggle switches and press buttons. I love pressing big red buttons and flipping switches! Why would I give that up?

    I love the comment above about so many IoT products just being a middleman between me and every on/off switch. That’s a keeper. If I can use it to convince someone not to hire me to develop a poorly thought out IoT ‘product’ then you’ll just have saved someone some money, and allowed me to spend my time doing something more productive, so thanks for that!

  11. IOT? I don’t need no stinkin IOT ? Why back in my day I used to walk to school 2 miles up hill both ways!
    Why I learned how to use a kettle back in Nam. We used to boil water in them to make tea, coffee and torture.

    Too soon? Perhaps the Nam joke was a little too soon. After all Kissinger is still with us.

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.