Helium token economics—complicated, confusing, brilliant?

Helium, the decentralized, block-chain based, network for the Internet of Things, seems to be off to the races. The company is selling Hotspots as fast as they can make them, and coverage has spread throughout much of the US. But the connection between how users pay for the network and how Hotspot owners get paid for providing the network is far from obvious. Here’s an explanation for your consideration. Be warned—it’s dense, and you’ll probably have to read it a few times. What it really needs is an infographic or—even better—an animation. But I’m not the guy for either.

Short Version

Users of the Helium network pay in Helium tokens (HNT).
Hotspots owners receive HNT.
You might think that the HNT flows from users to Helium, Inc., to Hotspot owners. This is wrong.
Rather, the HNT used to pay for network use doesn’t go anywhere at all. It is removed from circulation (i.e., “burned”).
Likewise, the HNT used to reward Hotspot owners doesn’t come from anywhere. It is created by the system (i.e., “minted”).
The rate at which HNT is minted does not change, but the rate at which HNT is burned will vary with (1) how much the network is used and (2) the value of HNT.
The system is designed so that, in theory, increased use of the Helium network will drive up the price of HNT on the open market.

Long Version

Helium uses a BME (burn-and-mint equilibrium) token model.
Helium tokens (HNT) are traded on open exchanges.
IoT device owners pay with data credits (DC) to transmit data via the Helium network.
DC may only be purchased with HNT. (Note, though, that for your convenience, Helium Inc. has set up a way for you to give them money via credit card, which they then use to buy HNT to buy DC to give to you.)
HNT that are used to purchase DC are “burned”; i.e., removed from circulation.
Data transmission via the Helium network can only be paid for with DC. (DC can also be used for a few other purposes, such as to pay the transaction fee for transferring HNT to someone else.)
DC may not be sold or transferred.
The price of data transmission is 1 DC per 24-byte data packet and, according to Helium, will never change.
The price of 1 DC is $0.00001, and, according to Helium, will never change.
Since the dollar value of HNT can fluctuate (because HNT is traded on open exchanges), the cost of data transmission in HNT (not in dollars) can also fluctuate.
Hotspot owners are rewarded for transmitting data by being granted newly minted HNT (and also for providing proof of coverage, but that’s not germane to this explanation).
Here is the key: The number of HNT granted to the entire population of participating Hotspot owners during a given time period (i.e., the amount of HNT “minted”) is fixed. It is not equal to or even proportional to the amount of data transmitted, and therefore it may or may not match the number of HNT burned to transmit the data.
The reward for a participating Hotspot owner for transmitting data is a percentage of the fixed number of HNT minted for that time period. The percentage of HNT received is proportional to the volume of data transmitted.
With the system at equilibrium, the number of HNT minted equals the number of HNT burned. However, if network usage increases, the number of HNT burned will exceed the (fixed) number of HNT minted, resulting in a net decrease in the number of HNT in existence. This, in theory, will cause the value of HNT on the open exchanges to increase. (Supply goes down, so price goes up.)
Because HNT is now worth more dollars, the number of HNT burned for subsequent data transmission (the cost of which, remember, is fixed in dollars) will decrease.
The dollar value of HNT will continue to increase, and therefore the number of HNT burned will continue to decrease, until the number of HNT burned is once again equal to the number minted. The system’s equilibrium is thus re-established as a result of the open market for HNT.
The dollar value of HNT on the open exchanges will remain at its increased level, provided that the usage of the network remains at its increased level.
Thus, in theory, the dollar value of HNT will be proportional to the amount of data transmitted through the Helium network. The more the Helium network is used, the more HNT will be worth.

PE Firm Acquires IoT Security Firm. Again.

The blanks in the following sentence have been filled in twice recently:

On [6th day of a month in 2020], [name of IoT security firm] announced it is being acquired by [name of private equity firm] for $[number>10^9].

Here they are:

On January 6, 2020, Armis announced it is being acquired by Insight Partners for $1.1B. (press release)

and

On February 6, 2020, Forescout announced it is being acquired by Advent International for $1.9B. (press release)

Just in case you needed convincing that there’s a huge demand for IoT security, there you go.

IoT and Hard Hats—What a Connected Construction Site Looks Like

The site is still nothing but an empty lot, but the groundbreaking is over, and the suits have gone home. Now the hard hats are arriving, and each one has a Spot-r from Triax clipped onto his or her work clothes somewhere. The system tracks worker location, detects falls, lets workers signal if they’re in duress, and notifies them in case of a site evacuation.

As the project begins and the activity ramps up, cameras installed around the site send video to the artificial intelligence algorithms of Indus.ai. The system recognizes the various construction vehicles and tracks their movement and activity. It records the arrival times of concrete trucks. It measures how long dirt haulers wait to be loaded. It counts how many scoops an excavator takes to fill a hauler. And it keeps an eye out for people without hard hats or vests.

As the concrete gets poured, it covers up the temperature sensors that have been attached to the rebar. They could be from Command Center or Concrete Sensors or SmartRock. The temperatures reported wirelessly by the devices in the hours and days after the pour can be used to derive the maturity of the concrete, which can in turn be used to derive its strength, which can be used to tweak and maybe even speed up the project schedule.

A temperature sensor set-up from Concrete Sensors

As the building goes up, sensor-packed devices from Pillar get attached to the structure. These monitor temperature, humidity, carbon monoxide, VOCs, particulates, noise, light, and pressure, and trigger alerts if something goes awry.

Pillar’s multi-sensor device. They call it a Pod.

With the interior work underway, a technician from Disperse comes on site once a week and uses a 360° camera to image every space. Disperse’s AI uses the images to compare the actual progress to the project schedule and to create status reports. It can also spot and prevent mistakes, such as putting up dry wall ~~before~~ when there’s still some plumbing or electrical work to do.

Here’s a marketing video from Disperse.

All the while, equipment is tracked and monitored via tags with sensors and wireless. The tags on the big, heavy stuff may be from Triax or Tenna, and there are lots of options like Digital Matter for the smaller equipment.

And once the building is complete, various, interconnected systems monitor water use, power use, air quality, and people’s comings and goings. But that’s another blog post.

Connected Bee Hives and Electric Kitchen Gadgets

Yesterday I ran across a company called Hostabee that makes a system for monitoring beehives. Today, I did a little research and ran across 7 more:

These companies sell systems comprised of a variety of sensors—hive temperature, hive humidity, hive noise level, hive weight, ambient light, ambient temperature, rainfall—and cellular or satellite connectivity. And they of course include backend databases, web portals, and mobile apps.

“Beekeeper at work” *by allispossible.org.uk is licensed under* CC BY-NC-SA 2.0

In other words, these systems apply the basic building blocks of IoT to yet another use case. Which makes me think of my kitchen, believe it or not.

I look around my kitchen and see a blender and a food processor and a coffee grinder, three devices that use an electric motor to spin a blade. I’m sure you can think of others.

I step out of the kitchen and see a circular saw, a weed whacker, a power drill, and a Dremel tool. Again, electric motors spinning things to cut things. Two building blocks satisfying numerous use cases.

Back in the kitchen, there’s a few things I don’t see. For instance, when I was a kid, we had one of these electric can openers:

I don’t have one of those. We also had an electric carving knife, and I don’t see those around anymore.

My point? We’re throwing IoT at all sorts of things right now. Some of it will stick, and some of it won’t. With the bees, I bet it will.

Who’s Gonna Track Your Wild Horses?

The use cases for IoT continue to amaze me. Hot on the heels of last week’s news of ship-detecting albatrosses comes this report of a system being used to track horses in Mongolia.

It seems that in Mongolia, people don’t like to keep their horses penned in and tied up–they just let them run free, and the horses routinely travel dozens of miles from home. Some horse owners there are keeping tabs on their assets as they roam the steppe by fitting them with collars equipped with GPS receivers and satellite transmitters. Most of these devices send updates every 4 hours, but valuable racehorses are wearing a version that transmits hourly, albeit at the cost of shorter battery life. And they don’t have to trick out every horse, just one in each herd of 30 or so. The devices, as well as the satellite system they transmit to, are by Globalstar.

“Mongolian Horses” *by Tinker Sailor Soldier Spy is licensed under* CC BY-NC-ND 2.0

Using Connected Albatrosses to Make Fishing Boats Follow the Rules

What do you get when you cross a radar detector, a GPS receiver, and a satellite antenna with an albatross?

No, seriously. That’s not a joke. A group of marine ecologists at the French National Centre for Scientific Research in Chizé strapped a device containing the fore-mentioned components onto the fore-mentioned bird (actually, a bunch of them) and set them loose to collect data about fishing boats. Albatrosses are great for this because they fly looooong distances and tend to be attracted to boats. When the device picked up a boat’s radar signal, it sent the location to the scientists via satellite.

An albatross (unconnected)
“Diomedeidae EM1B6915” *by Bengt Nyman is licensed under* CC BY 2.0

Why? International Maritime Organization regulations require that ships broadcast their identity, position, direction, and speed to other ships using an Automatic Identification System transponder, but sometimes they don’t, maybe because they don’t want the competition to know where they are, or maybe because they’re doing something they’re not supposed to. Those big birds with little backpacks picked up radar signals from 353 ships, but the scientists found AIS transponder data from only 253 of them.

This is the first time anybody has been able to collect this type of data, all thanks to the albatrosses.

My first thought when I read this was, wow, that’s cool. My second was, this worked once, sure, but from now on those scofflaw fisherman are going to shoot every last bird they see. The scientists claim that won’t happen, because the albatrosses keep their distance from ships, and because it’s hard to shoot straight from a boat while it’s bobbing up and down in the water. Hmmm, maybe.

(The results of the study were published here, and I also used this article and this one as sources for this post. I first heard about this when @gigastacey retweeted @ElasmoBro.)

How the NFL Could Do More to Protect Players’ Brains

A sports article popped up in my tech news stream last week reporting stats on a quarterback’s passes during a practice—how fast he was throwing the football and how fast it was rotating. Wow, I thought, how do they get the rotation data? A little digging and I learned that the NFL uses a system from Zebra Technologies to track the balls. It has 2 components: nickel-sized transmitters inside the balls and receiving devices placed around the field and stadium. The players have transmitters in their shoulder pads as well, one on each side, and their movements on the field are tracked and recorded during games. (More info here.) Impressive.

But they don’t put transmitters in the players’ helmets. Why not? It’s indisputable that football causes brain damage—NFL players develop cognitive impairments years after they’ve stopped playing, and their brains exhibit damage distinct from other neurological disorders. Remember Junior Seau? Remember Mike Webster? Couldn’t the tech help with this?

The league does have measures in place that attempt to mitigate brain injuries. Players who take hard hits to the head are assessed on the sideline and taken out of the game if they show signs of a concussion, but most of these signs—disorientation, a blank look, balance problems—are subjective. Why doesn’t the NFL put sensors in the helmets to measure the g forces that the hits produce and use that information when deciding whether a player should come out?

And they could do even more than that. Players who take big hits to the head but don’t show signs of concussion can subsequently be concussed by ordinary hits (described in detail here). With devices in helmets, the league could better protect players by taking them out when the helmet device transmits a big hit of, say, 80 gs or more.

So again, why not? It’s because the NFL wants to keep the players on the field. And so do the fans.

Smart Trails – Using Tech to Manage Hiking Trails

If you’re the steward of a system of hiking trails, how do you do your job? How do you decide how to allocate your limited resources? How do you get visibility into the use of the public resource that you’re responsible for? Can tech help?

“Culvers Gap Hike” *by Garden State Hiker is licensed under* CC BY 2.0

Last week, the executive director of my local makerspace connected me with Ryan Faulkner, project specialist with the Connecticut Trail Census. Their job is to measure trail traffic so that trail management can be informed by empirical measurements. To do so they’ve installed devices on trails throughout the state. The devices measure traffic by detecting the infrared radiation (aka heat) coming off people as they hike by.

Ryan and his colleagues currently must drive to the trails and hike to the devices to download their data. The devices, made by TRAFx, are spread out all over the state, so it takes time and money, which means they’re only able to do it quarterly. Ryan wanted to talk about how they might automate data retrieval so they could cut costs and get the data into their system more frequently.

We outlined his options:

Migrate to connected trail counters from another vendor, such as Eco-Counter.
Partner with a hardware developer to design and manufacture connected trail counters (and at the same time, power them with solar so that the team doesn’t have to replace batteries)

But if they developed a counter themselves, how would it connect?

LoRa—It’s a good low-power solution, but they’d need to install a cellular gateway within range of the trail device and find a way to power it.
Helium—Not an option yet, because the only gateway (the Hotspot) is not for outdoor use.
Cellular—This is the way to go, assuming there’s coverage at the trails and the device is affordable. All the better if there’s NB-IoT coverage.

While it’s fun to think about developing and manufacturing a device, and it would be fun to do, hardware development is no small thing. It’s expensive, risky, and time consuming. The good news is that they might not have to do it, because TRAFx is considering developing a cellular version of the trail counter that the Connectiuct Trail Census already uses. If that comes to fruition, all they’ll need is a hardware refresh and they’ll be effortlessly getting their data weekly.

What CBRS Is and Why You Should Care About It (Hint: Think 5G)

Yesterday the FCC authorized full commercial deployment in the 3.55 GHz to 3.70 GHz band, clearing the way for CBRS.

Huh? So?

That chunk of spectrum is the mid-band for 5G. CBRS is Citizen’s Broadband Radio Service, and the FCC wrote the rules so as to give access to the widest possible group of users.

This makes possible private 5G networks in a stadiums and campuses and sprawling factories.

It’s a milestone along the road to our connected future.

Your Body Was Doing Edge Computing Way Before It Was a Thing in IoT

I love cool technical facts from nature. Spider silk is stronger than steel! Prairie dogs use Bernoulli’s Principle to ventilate their burroughs! Bats use sonar to navigate and hunt!

So as I try to envision where edge computing is headed, I can’t help but think back to my neuroscience days and see parallels in biological neural systems.

Edge computing is, of course, um…wait a second. It’s not all that clear what edge computing is. The clearest definition I’ve seen is that edge computing is any computing that’s done outside of a data center. (I got that from Geoff Tate of Flex Logix; he says it at the beginning of this video.) So edge computing could take place in a sensing device or in a gateway or in a box at the base of a cell tower.

Now, if your brain is the equivalent of a data center, then any “computing” outside of your brain is edge computing, right? Here are some of my favorite examples:

Each of your eyes has about 125 million photoreceptors (that’s 10x the number of pixels in each camera in the iPhone 11, by the way), but your eye doesn’t send that pixellated information to your brain. It uses the processing power of four other types of cells, arranged in layers, to reduce that information to rudimentary yet salient features before shipping it off to the “data center” for further processing. (I have to point out that your retinas are not officially out on the edge but rather are actually considered part of your brain because of the way they develop. My neuroanatomy teacher, Larry Swanson, would be very disappointed in me if I didn’t.)
Sound is broken down into its frequency components by the very first cells in the auditory pathway. Your wetware computes a Fourier transform of the sounds entering your ears before “uploading” it to your brain.
Reflexes like kicking when the doctor hits your knee with that little hammer or when you pull your hand away from a hot stove are mediated entirely by your spinal cord. Data center (aka brain) totally unnecessary.
Walking. Cats don’t need their brains to walk, so you might not either. Have a look at this video. (Warning: It shows an experiment on a cat that the cat doesn’t survive, but it’s not at all gory, and it doesn’t show the death of the cat.)