Insteon's vanishing act explained: Smart home biz insolvent, sells off assets Smartlabs, Inc, parent of vanished internet of things vendor Insteon, is unable to meet its financial obligations and has assigned its assets to a financial services firm to be sold. After recently shutting down the servers supporting its smart home hub app and saying nothing to its customers or partners, the California-based …
Dear Insteon Partners,
It's been a great ride bringing you world leading internet connected cloudy thingy techno doodad, running on best-in-class ponzi financial engineering.
Unfortunately on Monday my mistress tried to order another crate of champers from room service, but we were shocked when the last company credit card was declined - apparently someone had been using it to pay for "hosting", whatever that is. Fortunately she had insisted on cash up front for the week, so my fun didn't come to an abrupt stop, unlike your thingies.
Rob Lilleness
"Which is why, you're going to do home automation, it should be local."
I can't remember a time when I wish I could remote control the electrical devices in my home from some distance away. If I remember I left a light on these days, it's LED and could be left on for months without a notable increase in my bill. If I leave on holiday, I make lists (yeah, I'm like that) and making sure things are shut off before the front door hits my backside is a task I do.
I do have some X10 gear so I can turn lights on/off remotely. I find that handy especially if I want to turn on an outside light if I hear something rather than getting up and padding to the wall switch on the other side of the house.
I do get that more complex automation is a benefit to the disabled, but why would that need to be connected to the cloud? There are systems that can be remotely access by somebody such as a family member or caregiver but don't need the internet to function locally.
Do the owners of the now not-so-smart hubs and devices count as creditors? I mean they did buy a service (the use of the hub), and they are no longer receiving said service. A few quid to allow them buy some replacement kit would seem appropriate, no?
Oh sorry, I forgot which universe I was in for a moment there...
"Do the owners of the now not-so-smart hubs and devices count as creditors? I mean they did buy a service (the use of the hub), and they are no longer receiving said service"
I didn't see if users paid a monthly fee. Many IoT junk shops include the online functionality when you buy the parts, but they seem to fold up as soon as sales slow down. Many/most people aren't going to be thinking of the long term viability of the company when they invest in the equipment, but they should.
There is absolutely no added value in the cloud server, it exists merely to justify subscriptions and to collect intelligence about the users. The industrial world has been handling much more complex automation tasks than the typical home user for decades and its never needed 24/7 Internet access, cloud servers and what-have-you. (...and its never needed constant firmware updates, either). The domestic world shouldn't need sophisticated back ends except for doing bleeding edge stuff like decoding natural speech and other environmental signals (the control itself is trivial).
Ultimately its a failure of the business model. Our industry can't compete by making things, it can only compete by buying imported (from China, usually) things and renting them to people.
You don't need cloud for that stuff, the whole thing can run off a 40 euro raspberry pi. The few thingies I have still talk to a raspberry pi 1B, The music stuff runs on a Pi 3. The only downside is that the new ones are always on backorder.
Industrial stuff is usually very expensive companies like Siemens buy a lot of the competition to force their own antiquated s7 and very user unfriendly tools.
Industrial stuff is also something you would absolutely NEVER want to grant internet access to, let alone anything outside and air-gapped or heavily firewalled vlan, primarily because most of the companies selling that stuff don't give a lick about network security and would rather sell you a forklift upgrade of their controllers for 6 or 7 digits which are only slightly less out of date than the massively out of date stuff you are currently running.
Oh, and their programming interface is also hideous- one of the systems out there uses Visio for it's workflow programming. ::retches::
"Industrial stuff is also something you would absolutely NEVER want to grant internet access to,"
Yep. Layer 2 NAT. It's like AWK - if L2NAT is the answer, you asked the question wrong... ;)
(I mean, I know it has its uses, but it is sometimes hard to get your head around, and even harder to remember long-term)
There is absolutely no added value in the cloud server, it exists merely to justify subscriptions and to collect intelligence about the users.
Actually it is really difficult to make a consumer level plug-n-go IoT device that has local and remote access without using a straight up remote server.
The "web" i.e. Chrome is inherently very hostile to direct connections, and you enter a morass of workarounds that they keep breaking in the name of security. Everyone who set web standards (google, and google, and , oh google) has/had a huge vested interest in driving the server only, cloudy model, and that's what they've made.
I totally understand why any IoT maker has little choice but to go that way. Even then you have to keep working on your product to keep it working in browsers, when it is beyond trivial. That means that it's almost inevitable that products will be abandoned and broken.
After 3 years of non stop breaking changes every 3 months, I reverted a product from using a browser interface to direct windows software. It has run with 0 (new) issues for 7 years. And it took a fraction of the time to implement.
> Everyone who set web standards (google, and google, and , oh google) has/had a huge vested interest in driving the server only, cloudy model, and that's what they've made.
I have no particular liking for IPv6, but putting my tinfoil hat on for a moment, I wonder if this explains why there’s really been no push by the industry as a whole to get IPv6 into the hands of consumers. Because a world where NAT is a choice rather than a necessity, a world where every device (heck, every atom of every device) can be made directly-accessible if required… is a world that won’t need to pay for pointless cloud-based IoT infrastructure.
Perhaps I’m overthinking it?
>> Actually it is really difficult to make a consumer level plug-n-go IoT device that has local and remote access >> without using a straight up remote server.
Not really... RadioThermostat does just this - you can leverage their cloudy based offering if you like, or access the devices locally through an API... which they publish documentation on and have received support on from them. Admittedly, they are not the nicest things to look at aesthetically, but it is possible and even can plug in either Wifi or Z-Wave radios.
If you replace "it is really difficult" with "most vendors are too lazy" then that is a FTFY.
Meaning of course belated fixes for all the stupid coding errors they left in the product.
Charging a subscription for such a 'service' is adding insult to injury. Particularly where IoT firmware elements are (as quite commonly) copied and pasted from chip manufacturers' data sheet demo code (ncluding hard coded credentials).
I've said it before and will keep on saying it: Until there is a properly designed and inherently secure specific IoT protocol which had been properly tested, is fully certified and has been adopted by the entire industry as the default standard then I'm not interested in this shit.That way, at least you should be device agnostic and it shouldn't matter if a vendor goes bust as you should still be able to use anything you've bought from them.
And yet thousands of people daily buy the shiny shiny expecting it to just work. It has an "i" in front of the name so it must be good.
In the beginning of the Internet, you needed to know what you were doing. Today, most people think Wifi is a single standard. The Simpsons would call it the endumbening.
A thousand monkeys with a thousand typewriters is more productive than a thousand people with "Smart" phones today.
I don't think you're being fair about endumbening. Simonlb's point above about standards is a good one IMO; the "internet", and it's various connection points/protocols/etc. is a utility for the vast majority of people out there. The other utilities (at least in the UK) are pretty well set up, standardized and backed up by law. If I buy an electrical product then it's BS/CE certified against well thought out and mature standards and I can safely plug it in and use it without having to understand discrimination, power factor, earthing, current consumption, etc. and even if I mess up the system's well designed so in the majority of cases there's a failsafe like a fuse or ELCB - even a shutter across the terminals in UK mains outlets. Electrical standards, building regs, consumer law, H&S, etc. work together to make the electrical products and the supply pretty safe and give consumers and the law redress against cowboys. We've all become endumbened about the mains supply cos we don't have to think about it.
Wouldn't it be great if the IoT were like that?
You mean you SHOULD be able to purchase it, take it home, and just plug it in without risking death or dismemberment. It turns out that some people, who should be publicly hanged, will try to sell us any old shit they can get for as cheap as possible from sketchy manufacturers, who should also be publicly hanged, that isn't even safe to simply buy, and then dump into a landfill, to say nothing about being fit for purpose.
Oh yes. There's a famous YouTube channel and associated website (http://www.bigclive.com) where Big Clive dissects 'wall-warts' bought from various sources, including online tat emporia and points out how many safety standards they break. I strongly recommend it. It is eye-opening.
Note that the ink to put an CE logo (even an accidentally non-compliant one) is a lot cheaper than actually doing the electrical compliance testing.
This applies even more to the post-Brexit landscape where manufacturers are less likely to go through the process to put a UKCA mark on devices.
"a properly designed and inherently secure specific IoT protocol which had been properly tested, is fully certified and has been adopted by the entire industry as the default standard"
It would be a strange world indeed where the internet of useless tat has better standards than the internet of banking, or the internet of motor vehicle registration, or the internet of you-name-it.
Anyway, there are more ways to fail than protocols and security, and IoT has found most of them. Failing to grow enough to pay for keeping the server turned on seems to me the most reliable point of unreliability.
"I've said it before and will keep on saying it:"
I'm an IT bod, an ex Civ Eng etc. I run a fair bit of my home and work on IoT. I apply the same approach to IoT as I do everything else. It has to work, with or without the internet, with or without wifi/Zwave/Zigbee etc. It will fail safe. I monitor my devices.
I'll grant you that there is a fairly high bar to entry but you are an El Reg reader so should be capable of at least finding your own arse with both hands.
My house is heated with electric underfloor mats. I've been testing a Zwave controller in two (out of 10) zones for three years now. One of them seems to go a bit mad every now and then but not the other. Anyway, I'll probably have a conclusion within about five more years.
This is new stuff. Do your due dil. and stop whining.
Damn, I hate whining.
If you absolutely want home automation and cams and controlling the home from the phone , lights , shades , whatever .. you can always go to the well established vendors like Crestron and AMX which have been in business for decades. I don't know how many systems i installed / programmed and the companies are reliable . You can do anything you want , go cheap if you want , but in the long run it pays to get good grade material from well established vendors. Gadgets will remain gadgets .. but real control systems are established technology and will give you decades of service. Buy a good one so you don't have to buy over and over again. I'm no longer in the business , im retired , so i got nothing to sell .. just good advice from an experienced man :)
Only the hub is IP enabled and it can still be leveraged locally without the cloud.
The real issue here is that it is a proprietary technology. Replacements are a long term problem but the lack of bridge devices are a critical failure point. Insteon is a mix of powerline tech (x10 compatible) and 900Mhz radios. If you have 80 devices and lose your bridge....you have 80 useless devices.
... I hope Universal Devices is able to do what they hope they can, and make that tech available.
But ultimately this is the problem with advanced things like that. I've burnt my fingers a few times with stuff like that (Hello being one, Lytro and 360fly being the others), so there is zero chance of me ever considering a 'smart home' or 'smart speaker' technology again, unless there's a path of alternatives (such as other controllers being compatible and able to reset devices).
I have a D-Link room monitor, from a few years ago. They've upgraded the firmware even after the product was no longer available for sale.
And they've announced withdrawal of the cloudy monitoring service a year before it gets shut off, in November 2022.
Hats off to them. It's how it should be done.
Still, "The Cloud"™ is _HIGHLY_ overrated...
A few years ago I was given one of those intarweb radios as a present for christmas or something like that. Cool idea, used it for years, then find out, no longer really works right any more, always dropping the connection and getting network errors [I can watch them with wireshark].all because Broadcom (who apparently had a closed source firmware thing going on all of these radios for a while, several different brands even) dropped ALL support a couple of years ago, and will NOT even allow the radio makers to SUPPORT IT THEMSELVES by UPDATING THE FIRMWARE on these devices and thereby keep the service going on their own... and therefore the radio effectively *DIED* because, hardcoded URLS (which I can observe in wireshark) and things like that. yeah, there are some hacks availoable to "make it work" but I do not want to do that. Making an RPi do its job instead.
So, if "Cloud" tech is SO FRAGILE and only lasts a HANDFUL of years before it effectively goes titstup, then UNLESS there is an OPEN STANDARD or OPEN SOURCE way of FIXING it ourselves, I'd say:
DO! NOT! BUY! CLOUD! TECH!!!
And THIS hurts their entire business model, not just because of me, but the bad taste that MANY OTHERS get in their mouths over the UNRELIABILITY of future support for something that SHOULD last for DECADES.
> The notice explains that in 2017, Smartlabs had financial problems and obtained additional capital and new management to improve its situation. But the emergence of the global pandemic in 2020 and the ensuing supply chain disruption proved too much new management turned-out to be not nearly as good as they thought they were, prompting the company to seek a buyer in November 2021.
FTFY
With the 'global pandemic' that's been going on for the last two years, I would have expected this business to be booming. You have 80% of the population stuck at home and lazy as can be, and you can't even sell gizmo's to keep Joe Schmo from having to get off of the couch to turn off a light? Bullshit.
If one sells a device to consumers, and included with the device is a promise to supply certain server services to the owner of it in perpetuity, obviously we need laws by means of which that promise is enforced.
If the servers shut down, the government seizes them, and uses the IRS to bill the company for keeping them running. And, of course, the government has absolute priority over all other creditors.
If companies know that's how things like this will be handled - they will invest part of the price of such products into something that will yield enough interest to pay for the servers.
But I'm afraid that isn't realistic, if only because there are no ironclad investments that are guaranteed not to lose money and to have a minimum yield of interest that can be relied upon. Currencies inflate. There's even a movement afoot to abolish paper money so that negative interest rates can be brought in.
What I still haven't figured out is how we can abolish inflation by going back to the gold standard, and yet have fiscal and monetary policies that ensure full employment forever.
Perhaps a better take on that: the requirement to publish the specification so that competing services can support those devices (including non-profits and home-grown).
So if 'example' company goes titsup and the 'example.com' web service is no longer available, hackers should be able to produce an equivalent service, upload to "git repo someplace", maybe ask donations like with phone apps and firefox extensions, and then you do a config change on your device to point to "example.org" instead (or whatever, maybe even an IP address on your LAN) and then you get the functionality back (or a good portion of it, depending). And ISPs could even help by making it possible to tunnel and get a listening port for things _LIKE_ this.
That's how I see it, anyway. And the requirement to publish, and make firmware updates available (or source so that they can be MODIFIED) could be an "escrow" requirement before being allowed to sell the product, so that it only kicks in if the online service being provided is in any way discontinued. Then, customers have the opportunity for 'relief', which is better than having to throw "that useless brick" away, now that the company that manufactured and/or sold it dies.
Buying a cloud product has AN IMPLIED CONTRACT. It should be ENFORCED that way. At least, in "common sense" world...
"Perhaps a better take on that: the requirement to publish the specification so that competing services can support those devices (including non-profits and home-grown)."
Publishing the spec would be every so helpful, but negating any Copyright, Patents or rights of DMCA takedowns could be helpful as well if the company doesn't have an immediate successor. If there is the possibility of somebody offering a replacement service they can make money with, the gap will be filled.
Let's say that a few of the hundreds of lawsuits currently active against Tesla and Elon Musk go against them and the company folds. In the US, Tesla vehicles use a proprietary plug for DC fast charging and a link with the car to bill the customer for power as there is no payment functionality at the charging point. They also seem to rush their software out before it's fully tested and need to send out frequent updates. If fast charging the car is limited because the Supercharger stations are shut, the app needed to unlock the doors goes offline and updates to critical safety functions are no longer being updated, that could be a huge problem for a product that's so expensive. The impact is not just to the customer but to banks that lent the money for the purchase. We all know that government sees banks as Too big to fail so all those banks would clamor to be bailed out or Tesla would have to be given a free pass if they break the law (again).
"If the servers shut down, the government seizes them, and uses the IRS to bill the company for keeping them running."
You are really advocating that the US government gets into tech? When they took over a brothel in Nevada they couldn't make a go of selling booze and short term relationships.
People need to look at who they're buying from and make the right decision.
Yes, Amazon, Google et al shut down niche or periphery services on a regular basis, but they're the only people I'll buy any smart devices from.
This is because those devices are so intertwined in their eco systems that to suddenly kill them off would be a significant undertaking with a very large, angry mob.
It's harder to use herd protection when there's only 100000 of you.
But then you're putting longevity before massive private data slurping.
The only good IoT thing is one you make yourself.
It's relatively easy (for an IT person) to use something like an rpi zero w as a hub/thermostat (using Tinycore so read only), and then use something like ESP8266 devices with decent temperature sensors (ie MCP9808 or better yet BME680) to make far better IoT/home automation sensors than the expensive (pretty) commercial rubbish people are buying.
you defined some actual good product ideas.
Now, instead of making them contact an outside server, run something on a PC or your phone to contact the device. Make it open source so you sell the hardware, and the software supports it as long as there are people willing to maintain it (including customers). This becomes a selling feature.
"It's relatively easy (for an IT person) to use something like an rpi zero w as a hub/thermostat (using Tinycore so read only), and then use something like ESP8266 devices with decent temperature sensors (ie MCP9808 or better yet BME680) to make far better IoT/home automation sensors than the expensive (pretty) commercial rubbish people are buying."
I was going to use an Arduino set up to control a solar heating system I installed on the house but found it much easier to build a dedicated bit of electronics to do the job that is much more robust. If/Then sorts of logic is dead simple to do in an analog way. I'm more of a hardware person anyway. I thought the project would be a good exercise in coding, but in the end decided to save myself time and look for a project where having a microcontroller is more necessary.
A couple of things still don't quite add up. If we take the story at face value for the sake of argument you still need to explain
1) Why the company took this long to let customers know it was going under, in fact actually going under before bothering to tell anyone and leaving customers speculating about what happened
2) Why the company execs started trying to scrub all traces of their involvement with the company ahead of announcing the company was going under
While I don't doubt that supply chain shocks put a number of nails in the coffin of the company, the other actions of the management suggests some serious blunders (best case) or outright fraud (seems more likely given the attempt to scrub all association from LinkedIn).
I hope Rob Lilleness's vanishing act is widely publicized. Company fail to make enough money to stay afloat? It happens. Need to file for bankruptcy or equivalent and shut down services? Give your users some warning, for heaven's sake. But pretend you have nothing to do with the whole fiasco? Unforgivable. I hope the word gets around. This guy deserves to be unemployed for a long, long time. After the way he's acted, I wouldn't hire Rob Lilleness as a janitor.
"Company fail to make enough money to stay afloat? It happens."
In this case the writing was on the wall for an extended period of time. I'd love to see the books. I see so many companies that are incredibly top heavy where the C-Level compensation plans are the problem.
I was looking at a company where an old work mate went to and their website shows a long parade of suits "managing" the company but nearly no mention of people that do actual work. I know a couple of the corporate members and they are the most useless people I have ever met other than their talent for accumulating money.
As an IT website, the focus here is the reporting of computer based shenanigans including the pathetic abomination referred to as IoT. Not sure how Insteon became entangled into the (not so)SmartHomes cloudy endeavor but it was a very good stand alone product for many years. Similar to power line networking (but preceding it, I think) It used the power lines to communicate between modules. This enabled switching and controlling lights/receptacles, etc. without running gobs of wiring everywhere (and tearing up walls when adding things after the fact). Other than local control, if desired here could be a master controller plugged in somewhere. Insteon was a great improvement over the similar, but somewhat buggy, X-10 system. Insteon hardware solved thorny problems for many of my clients and I, for one, am saddened to see it going down in flames.
"without running gobs of wiring everywhere"
Why would common Wifi/Zigbee/RF devices require gobs of wiring everywhere?
Besides, it's not their technology that's under fire, it's the cowardliness of their upper management and shameful communication with their customers.
"Why would common Wifi/Zigbee/RF devices require gobs of wiring everywhere?" You're confusing wireless devices with a control system using existing electrical wiring. Gobs of wiring means running switch legs to the new locations and/or running new wires between switches if controlling lights (or plugs) from more than one location). Of course, if adding a new light you have to at least get power to it by wire but even then, switching that light can get complicated. Been there, done that 1000 times over.
"Why would common Wifi/Zigbee/RF devices require gobs of wiring everywhere?"
They won't, but they might also not work. Depending on the area the devices are in, RF communication isn't always a feasible option. This can be the case for many types of environment.
Perhaps the device and the central controller are spaced widely apart if you have a large house. In order for a signal to make it across, you may need to create a mesh network out of the stuff in between. Since that's not going to happen unless you already have the skill to DIY a lot of this, it becomes a problem. Maybe you've got a house that doesn't allow signals to penetrate the walls. Some kinds of buildings really attenuate signals such that even using WiFi through a wall doesn't work consistently.
If you live in a more densely-inhabited area, you may also have interference problems. Every area where people live closely together has a ton of 2.4 GHz networks that are clashing to the extent that 5 GHz or the new band on WiFi 6 that not many people have yet are the only feasible options. IoT gear using WiFi tends to only support 2.4 GHz frequencies using low-powered WiFi chips that already don't have great range. Put them in a place where your neighbor's network is stronger than yours and they're likely to have spotty connection.
There's a reason that a wired connection can often be more reliable. Adding more frequency area can help, but not in all situations.
I've been burned by a few 'smart' home things that more or less have all decided to stop working at random:
Pretty much anything using a ESP32 just won't really work out well with the wifi I 've got. There's a configuration switch buried somewhere in Ubiquiti's management app to make it play nicer with 2.4Ghz devices, but I haven't really found the time or the willpower to dig it out.
I have about a half dozen of the smart plugs by three or four manufacturers- they've all done loss of control on me, and the only way to get them back is to utterly remove them from the controlling system, factory reset them, and re-add them. Same thing with the nearly dozen Phillips "WiZ" wifi lights.
After the third or fourth time of having to re-build everything from the lighting configuration to the schedules and cursing the entire time, I went Zigbee/Zwave with everything- all the lights and the smart plugs are Zigbee based and homed to a Hubitat box. Why Hubitat? because it has a built-in hook to talk to the echo dots for voice control, which are the only things that haven't broken in strange and stupid ways. (even the amazon smart plugs crapped the bed on me.)
The only regret I have is that the WiZ remotes were actually Not Bad, but since they are wifi only and no one has no information on how they communicate, they are in the E-waste box along with all the other wifi crap. (There are some Z-Wave and Zigbee button controllers out there that seem to work pretty well; There's also a third-party add on for Hubitat that leverages a Heroku instance to communicate back to the Amazon echo devices, so I can push a button next to the laundry room and a timer in my office kicks off automatically.)
My next goal for localizing my smart home is getting rid of the amazon bugs echo devices and replace them with on-prem voice control boxes, either powered by multiple raspberry pi units (and a spiffy microphone/LED HAT) with bits of glue code, or USFF windows boxes with Vox Commando and some glue code to make it play with the Hubitat.
The nearest devices I have to IoT are some recent model cameras and lens from well known manufacturers. Every so often there are updates issued by these manufacturers which are DOWNLOADED to a computer then manually installed to the device. A bit old school but the updates are infrequent. More importantly, if any of the manufacturers does die the gear will still work as none are dependent on talking to server.
If you can't host 100 million home automation devices on few ordinary servers, you're seriously doing it wrong. Remember that X number of clients also means X free computational nodes. The servers should only coordinate communications and hold configuration backups.
I imagine all these makers that cut hosting the moment budgets get tight have been putting too much faith in cloud vendor blogs.
Insteon has been around for a looonnng time, since well before anyone thought of IoT. I'm sorry that they have gone under, the loss of their server for the Insteon Hub is only a minor impediment. Most people who have used Insteon devices have also used one of the non-web dependent hubs or computer software, though the devices were able to communicate and control each other without that, and certainly without the web (though it is nice to have that available to use VPN back to an intelligent hub/controller).
It is a powerline technology, like X10 was, but they also later implemented radio link as well, and unlike X10, had 2-way communication, instead of merely hoping that a command had been received. Most home automation wireless systems piggyback on the wifi bands, where wifi, bluetooth, microwave ovens, cordless phones, and too many other devices stomp all over each other. I forget what frequency Insteon used, but it isn't the wifi bands as I recall.
Hopefully the new owner will continue production and/or release sufficient information for others to produce the devices (Insteon only allowed a few other companies to produce devices, and I think then only with modules purchased from them). Configuration can be complex, but real applications of anything often are. So there is a lot to criticize them for, but IoT isn't really it.
My kids think I'm an old fogey 'cos...
I have my hard copy CDs reipped to my own MP3s on my own phone ( ditto DVDs on media server )
I have my home automation in house only and in a limited fashion that can be manually overriden any time.
I have a bog standard petrol car with limited gizmos and gadgets.
I never buy video game subscriptions to rent games monthly, buy my games outright on Steam ( I can survive if Steam ever folded )
I don't mind subscriptions and I don't mind modern gadgets, but don't sell me a bullshit picture of your ass and then tell me it's high class porn!
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
