IoT is a new name for an old game. It promises Heaven on Earth but could go to Hell if not done well.

There is a growing buzz about 5G and the implications for IoT. I discuss where IoT seems to be today and will speculate about the future.

IoT promises to enhance our lives. But the potential security flaws could become nightmares and the stuff that science fiction horror movies are made of. An American Casino was recently hacked through a “smart” internet thermostat control device, and their database of high-stakes gamblers was compromised. There are also very scary and grave consequences about personal privacy which Europe and some of Asia are positioned to handle much better than North America.

I have played in the IoT field before and after @Kevin Ashton coined the phrase “Internet-Of-Things” way back in 1999, but kudos to Kevin for thinking of it. IoT is more than a buzzword. But as with other “up and coming” technologies like blockchain, there is a lot more promise and hype about the future than there seem to be working solutions today. In a world where almost any solution you can imagine can be commercially deployed in less than a year, why is IoT “not here yet”?

The short answer is that IoT is here already. It has been here since before the “internet” was born. The internet merely expanded the network infrastructure and standards needed to enable more IoT solutions. But the real reason we are “waiting for the future” of IoT is because of the networks needed to enable the full vision of future solutions. The future will see orders of magnitude more IoT devices than the number of smartphones, PC’s and TV’s combined – many more devices than people on earth (plus dogs and cats). In the future, we won’t even realize how many IoT devices are embedded in our homes, cars, and even lurking in our pets and our own bodies.

The IoT Business Model

Today, IoT is mostly a vertical market / enterprise solution space, not yet as big in the consumer solution space but getting there. There are IoT devices for your house and car – like doorbell cameras or dashboard cameras that upload video -- that consumers can buy and install themselves and enjoy without an IoT service provider in the middle, but the vast majority of future IoT solutions will involve a managed service provider and ongoing fees. And traditional Mobile Network Operators are hoping to grab the network side of the business whether using specialized networks or cellular – pay as you go for packets. Some say that “nothing is free” although we typically imagine that the internet is free, but many companies hope to build specialized “toll road” networks for IoT that allegedly offer better security than the open internet.

Where did IoT come from?

As with all “things” on the internet, there is a client and a server, even if the server is a cloud. The client collects data, sometimes crunches the data before sending to the server, and sometimes the client receives data and instructions back from the server. Yes, there are some peer-to-peer IoT solutions possible, but for the most part, IoT is about a remote sensor reporting its data back to a server.

In the ancient days before the internet, back when dial-up modems and dinosaurs walked the earth, there were other wireless networks in use that tied remote sensors to a server. DataTac. Mobitex. CDPD. Metricom. Even dialup-modems over old analog cellular networks were used.

One of the largest package delivery services still thriving today pioneered delivery confirmations using barcode scanners, a resistive touchscreen to capture a signature, and tied all that to an optimized analog cellular fax modem to transmit short 15 to 30 second data streams with an image of the signature and the delivery timestamp over analog cellular. Although this predates the “internet” as we know it today, that solution was one of many that today would qualify as “IoT”.

Other examples of old and still existing solutions are inventory control sensors on vending machines and point of sale terminals, networked gas and water consumption meters, “smart” (and demonic) parking meters, alarms of many kinds, not to mention a slew of vehicle telematics sensors. In fact, even the original RFID devices, even barcodes, were actually sensors containing data where a handheld scanner brokered the transaction to harvest the data and relay it back to a server. Symbol Technologies is an example of an old company specializing in sensor scanner devices that arguably was an IoT enabler before the internet was a thing. So the concept of IoT is very old and just getting better over time.

Current Solutions

Your smartphone is an IoT device. This is especially true when using the “find my iPhone” feature to locate your lost device (or figure out where my teenage daughter is hiding). But other examples besides the ones mentioned above include refrigerators that email you when your milk is sour, a vast ecosystem of vehicle telematics solutions for connected cars, the growing number of shared-use rental bicycles and electric scooters with tracking devices tied to a mobile app, remote sensors for home environment control, and wearable sensors. There are hundreds of existing IoT solutions today, mostly flowing over internet protocol networks, which include TCP-IP as well as “send it and forget it” UDP packets.

There are many wearable sensors like Fitbits and Apple watches that are headed towards IoT with many already there. My niece was recently saved by her Apple watch indicating that her heart rate was spiking, which she ignored until her watch kept alerting her, and then she realized she was having an allergic reaction to antibiotics she has just consumed. We bought my elderly mother an Apple watch as a safety precaution in case she falls at home – provided she’s conscious, she can call or text for help. Lots of similar such devices exist today. But there are some interesting, shocking, and creepy wearable and internal sensors coming soon for people which are discussed later. The key to most IoT devices and applications are that they are “automated” to perform specific functions.

The Network is King

The network determines the solution capabilities. The network dictates client-server data communication in terms of latency, reliability, geography, capacity, data verbosity to and fro, and perhaps most importantly, data cost. 5G promises to reduce latency and increase network capacity and device density to enable a crowded freeway of self-driving cars as well as AR and VR solutions. But there are already IoT network solutions available today. Even now, AT&T and Verizon offer IoT data solutions for a low monthly fee based on data packet usage tiers – asset management, inventory control, telematics, smart cities, and mobile commerce.

5G will only be as good as its deployment footprint, which will not be everywhere. There is a strong potential for alternate, parallel, walled garden IoT networks for certain use cases. Some IoT applications don’t need realtime network coverage. Others do.

If you design your IoT system to tolerate network connectivity intermittency, and store precious data in memory while on the dark side of the moon, and add a CPU with logic to analyze data and make decisions autonomously, then… you can succeed while out of network coverage, but your system complexity goes up. Like connected cars on a crowded freeway that pass through dead zones in network coverage, onboard sensors and autonomous data analysis and decision making will make or break realization. (Speaking of connected cars, I wonder if they will brake for dogs and cats running across the street? What about a deer?) Cellular network coverage dead zones exist around my house and neighborhood. I cannot say how often it happens for satellite. But surely no network is fault proof.

5G promises to become the IoT roadway with all the exits and on-ramps and gas stations needed to build the future whiz-bang IoT system, but you choose your vehicle and destination and payload to build a business and make it relevant.

SigFox and LoRa Examples

One of the most intriguing IoT solution enablers is SigFox. They invented an amazing network and modem technology. A SigFox modem is supposed to be very cheap, under $10, predicted to cost about the same price as a fancy Starbucks drink in the future. Although it supports both send and receive, it is probably best used as a one-way, send-only sensor module. Once per day, at a random time, the modem bursts out a few-hundred bytes of data which will hopefully be caught by the network. The network can be a single antenna receiver with a potential range of over 25 miles in radius if appropriately stationed. The number of sensor-modems in the field probably has a hypothetical limit but could number over a million. The sensor-modem RF technology allows you to bury the modem even under a couple feet of dirt, and the battery lasts for years. Imagine a field full of crickets all chirping on a summer night. The network and system detect and decipher each individual cricket’s chirp. Each cricket has something to say, and your future IoT business model depends on hearing each individual cricket’s message.

What kind of solution would this enable? The sky is literally the limit because theoretically, it is possible to use an overhead satellite to catch the data bursts instead of a terrestrial antenna. So for situations where a traditional Mobile Network Operator cannot and will not deploy cellular coverage, this could be a solution. Whatever remote sensor you can imagine that only needs a short data burst to report status once per day would be perfect.

SigFox owns their proprietary technology. They invented the modem technology but licensed the design to hardware partners to build the modems and drive down cost. They wish that big, traditional Mobile Network Operators would buy into their technology, deploy and maintain the network, and pay fees to SigFox. It’s up to you to hook up a data capture sensor to the modem and start your business. SigFox is big in Europe but not so big in North America. Parts of Asia are jumping on the bandwagon.

One of SigFox’s competitors is LoRa, an industry consortium of different companies cooperating to create an industry standard, open specification. They call themselves “LoRa Alliance”. I like the name because it reminds me of Star Wars’ “Rebel Alliance” and I imagine what kind of logo they should all get tattooed on their arms – maybe a tattoo that turns each human member into an IoT device! But the “rebel” part fades away when you see the growing list of partners: Cisco, IBM, ZTE, Foxconn, NEC, Fujitsu, Orange, Bouygues, SK Telecom and hundreds of others.

LoRa’s network model is more full-duplex, send and receive, than SigFox’s. Both have pros and cons. A good technical description of SigFox versus LoRa is here: https://www.link-labs.com/blog/sigfox-vs-lora

And to be fair, there are already existing networks as part of cellular. So if you have an IoT idea, you can start now.

Here’s a good technical write up about cellular IoT networks: LTE-M versus NB versus 5G: https://www.leverege.com/blogpost/cellular-iot-explained-nb-iot-vs-lte-m

Security

On the security side, there are multiple issues. One type of hacking is losing access of your private data to an intruder, like your credit card information, which could lead to fraud. Certainly, if we have more and more IoT sensors and modules gathering data and invisibly sending that data into the cloud, we hope it doesn’t fall into enemy hands. But the other type of hacking could be far more serious, when a hacker breaches your security and can actually take over your system. This second type of hacking can potentially end the world as we know it.

Imagine a hacker releasing a virus into self-driving cars. Mass accidents. Or imagine a high-tech assassin just takes over your car and careens it off a cliff. (Expect to see that in the next James Bond movie.) Or all the networked voting machines in an election return false votes? Or IoT home garden sprinklers open their valves all day and flood? Or your home appliances get hijacked to all “turn on” all day and spin your utility meter faster than a Las Vegas slot-machine, but more expensive.

Impossible, you say? Not as far-fetched as you might think. Lookup “Stuxnet” on the internet or just read this article: https://www.forbes.com/2010/10/06/iran-nuclear-computer-technology-security-stuxnet-worm.html

It has been widely reported that some newer model vehicles with keyless remote entry and start features have been hacked and subsequently stolen by savvy thieves. There already are front porch security cameras that stream video to your smartphone when the doorbell rings. Tie that to a keyless entry system to let the Amazon delivery man put your package in your entry way and voila, a beautiful target for hackers.

So how do we proceed? Security is never fool proof. But security gets better and better and tries to stay one step ahead of the bad guys. Perhaps there is a blockchain security solution for all of these small IoT modules coming soon?

Learning from the Past

Here is one of my favorite stories about the potential value of IoT… A long time ago, an eternity in dog years, I met a very clever man who worked in a semiconductor foundry. Semiconductor factories and PC board assembly plants use a lot of specialized machinery, usually arranged in assembly lines, with as much automation as possible. The machines are very expensive. In this guy’s specific story, his foundry ran 24/7/365 and could not crank out product fast enough to meet demand. It was like printing money. Business was booming and so lucrative that his company installed redundant assembly lines in the foundry to insure that no single machine breakdown would halt production. We’ve all heard of “triple-redundancy”, a-la NASA, but at that time, they had 9-fold redundancy production lines. The foundry output was that important.

His job started as building and facilities maintenance, but he was much smarter than that. He noticed that the production line equipment was made by many different manufacturers from many different countries. Each specialized piece of machinery came with a user manual and a warranty, and usually also had a maintenance and support contract for onsite repairs – just like most enterprise office printers and copiers. But the problem was that his company did not want the foundry line to halt even for a second. The moment any machine broke down, they just failed over to another assembly line, they didn’t call a repairman to schedule an appointment.

This smart guy decided to actually read every manual for every piece of gear, and he discovered that most of them provided specifications for operating temperature, certain fluid levels, and even suggested preventative maintenance based on usage. He started slapping sensors on every machine. Temperature sensors as well as accelerometers to measure vibration, and he began gathering and recording operating data over time. After a few months, he figured out which machines were about to keel over and die, and he could predict when to call the maintenance guy, when to swap out a piece of gear, and when to fail over to a redundant assembly line.

His methodology enabled his company to reduce their redundancy from 9 parallel lines down to 3, saving them over $1Billion in cost at his foundry alone. He was a hero.

This is an IoT example not only of remote sensors tied to a network reporting one-way data, but also of Big Data analysis. Anybody can buy a database and set it up, that’s just technology. But figuring out what sensors to feed the database, how many sensors, where to put them, and crunching the data to predict the future outcome is still an art. Potential IoT solutions could be goldmines.

The Dark Side

Question: If a manufacturer implants a sensor in something you buy, and it doesn’t send or receive data over any network that you control or pay for, then is there a privacy issue? It’s like the old philosophy question: If a tree falls in the forest but nobody is there to hear, does it make any noise?

Certainly, there already have been sensors in cars for a very long time related to the “engine computer” of which you, the owner, know little to nothing about. When you take your car in for maintenance and a tune up, the mechanic often “reads the sensor” attached to your engine computer to pull “diagnostic codes” indicating what may be wrong or about to go wrong with your car. Suggested maintenance follows from those embedded sensor readings.

But what if future car sensors started recording driving habits, like how fast you drive, how high the engine RPM’s spike, how often you actually change your oil, and what if any potential future buyer of your car could access those records and decide if you were a good car owner or a bad car owner? In that case, the good versus evil decision depends upon which side of the equation you sit.

What if your insurance company or the police had access to your driving habits?

Several smartphone OEMs have already contemplated installing IoT sensors and modems in their smartphones that won’t send or receive data over your wifi or cellular network, but over an alternate, specialized IoT network. Why? Allegedly to monitor battery and electronic component lifecycles, perhaps even provide “free” device location lookup if lost. But the potential for abuse is there. Why not also capture your usage behavior?

How about an IoT sensor in your television? Your PC? Your home WiFi router? Google and Facebook have already turned your PC into a remote sensor to track your usage behavior, but a separate sensor in your laptop or smartphone that you don’t know exists could signal your presence to a TV at the sports bar to deliver targeted ads to you even in a public place, or just alert Big Brother that it’s you sitting in the living room watching TV alone without your roommates or family members.

The Apple Watch has some growing health monitoring capabilities that have made the news because some health care insurance companies are subsidizing the purchase of Apple Watches for their customers, apparently hoping that a wearable sensor will promote healthier lifestyles as well as warn about potential heart issues.

But do you really want your personal health diagnostics to go directly to your insurance company? Your driving habits from your car? Your usage behavior of your phone, TV, and PC back to the manufacturer? Europe and Japan already have stricter privacy laws than America regarding usage behavior tracking, which is why some websites of European origin have been asking your permission to allow cookies to “help you” with your browsing experience. Their privacy laws require you to opt-in before any cookies can be dropped or usage data can be captured. Those liberal Europeans are even contemplating using blockchain systems for end users to “own and control” their personal usage behavior data, and restrict access to companies wishing to track them. How Un-American!

Some new IoT usage models may deserve scrutiny and warrant regulation.

The Future is Bright

All of the traditional IoT solutions are just getting better and expanding. Remote sensors of all kinds are proliferating. But a helluva lot more are coming, especially wearables, security, home health, connected cars, and remote eCommerce.

I just paid $79 for a house call from an appliance repairman. My dishwasher is broken. I used several YouTube videos to help me self-diagnose what’s broken. But the appliance repair company had to send a live human to see for himself. Sure enough, I was right. Did the technician bring the part? Nope. Another service call is needed for the repair. I hope and pray that future appliances will just send diagnostics to the shop so they can show up with the needed parts in just one home visit.

RFID tags injected into your pet’s skin to help find lost pets is old news. Now imagine a chip embedded in your hand that replaces your key-fob and your credit cards: just wave your hand to open your house door, start your car, open the secure doors at work, turn on your tv and control the volume, change channels, and buy your drink at Starbucks.

Soon, people will be walking IoT devices. There are already companies and devices available to perform these functions, if you’re willing to swallow the Red pill.

Today there are several companies already offering or soon to offer “internal” devices such as subcutaneous devices embedded in your body. (One such device is the size of a credit card and sits under your skin. It glows when connected to Bluetooth. Stick it under the right tattoo as a party trick and you’re bound to get free drinks!) Grindhouse Wetware already has such devices that use Bluetooth to pull your biometric data, such as body temperature. DuoSkin is an evolving temporary, metallic “tattoo” that adheres to your skin and allows you to control external devices using gestures. Another example is OmniTouch, a “wearable computer”. Note: these are consumer devices, not hospital devices. It all reminds me of that character in The Empire Strikes Back, Lando’s aide, the bald guy with the computer mounted to his skull.

I’m not quite ready for an IoT device bolted to my body, but I am interested in some vehicle telematics that message my mobile phone with reminders about oil change, fluid levels, tire issues, as well as when my wife and kids drain the gas tank and leave it on fumes in the driveway the night before I have an important early morning business meeting at the office.

My mother and my mother-in-law are advancing in age. They want to remain independent and do not want to be forced into an assisted living retirement complex. An aging population will need multiple IoT devices to help the elderly stay safe and intervene when needed. A Godsend will be safe, reliable, self-driving cars to take them to the store and to their doctor appointments. And another Godsend will be devices that alert family members or medical professionals when a sudden health issue is in progress. Who knows? Maybe your service dog will be rigged up with all kinds of health sensors to help their master? Every time you pet your dog or he licks your face, embedded sensors monitor your health! In that future, I’d name my dog “Sparky” for a totally new reason.