Helium isn’t truly world wide. Swarm is more like a direct competitor to Iridium Short Burst Data (SBD) service. Works anywhere on the globe. Helium relies on people nearby running a node so it only works where people live and have an internet connection.
I love this type of tech and want to get in on that sweet 144,000 bytes of data at 5 per month, but... it's not just 5 per month, it's a minimum investment of 500 unfortunately.
Also, not sure where to find a decent Commodore system these days to utilize this plan :)
Sure, but 40 years have passed since then and 144K peanuts for any real world application.
A quick example, you have a sensor that sends a single metric every minute. Suppose that you can pack that info (sensor id, metric type and data) in 32 bytes. You're looking at about 1.2MB/month of data (8 times your 144K), which will come to about $40/month.
Quite expensive for a single data point, but I can imagine several use cases where that cost can be easily justified.
For use cases of this product you will be packing much tighter than this.
You get terminal identification from the carrier, no need to waste precious packet space on identifiers. And it's probably also a waste to encode field names into your packet - just record fields at the same offset every time. If 16 bits of precision is enough for you (it likely is) you can squeeze 16 metrics (or 16 time series recordings of the same metric) into every packet.
Up to the minute sensor data for something so remote that it needs a satellite uplink is either going to be for something incredibly critical or totally overkill and a waste of money. With Swarm, you could be waiting up to 2 hours between passes that last anywhere between 10 and 50 minutes so it's not like you'll have your data instantly for large portions of the day.
Yes, but then IoT usually invokes a sense of (almost) real-time monitoring of things and lots of packets etc...
But I get your point, also you can always do a lot of things to save on bandwidth, like do not send info which is not interesting (i.e. no point in sending 100s of "no fire was detected" messages).
Btw, anyone here knows if there's compression schemes designed specifically for small data packets? Gzip and others would be overkill as headers vastly exceed payload size. Just using raw LZ77 may work but it's 2022 so there's probably a specific thing for that already.
Also, what about data that follows a specific format, like only integer numbers, it would be nice to have an algorithm that takes a "string" of 32-bit ints and gives you back a binary buffer with a smaller lossless representation of it.
This product is targeted at actual users of "IoT", think remote monitoring in remote locations. Not "smart microwave" style IoT.
Not sure there is any automatic solution for compression here. If you know your own use case you are in the best position to choose where to sacrifice accuracy with a lossy compression scheme. But this relies on understanding of the accuracy of the input data, possible ranges of values, importance of accuracy in different fields etc. Not sure how an automated algorithm could take these real world constraints into account.
A clever compression algo can't help you if you try compress 4 doubles, but in reality you only needed byte precision on some fixed 0-100% range.
Yeah and that's probably why IoT devs absolutely crank up the data rate and spew dozens of Bluetooth advertising packets in the air for no apparent reason, probably significantly reducing battery life that could have been 5yrs, for applications that really don't need instant response.
Something like msgpack might be able to compress ints to some degree since it can represent them as smaller data types.
> Suppose that you can pack that info (sensor id, metric type and data) in 32 bytes
I suppose you haven't worked with such constraints before, so that's why you think this needs 32 bytes. In reality, 4 or 6 will likely do.
Send diff in metric value if applicable, use variable-length encoding to not waste bytes. Allocate bits and not whole ints to things (like metric type). ID is already part of lower level protocol - the address, so you do not need it
How do you use variable-length encoding in a limited-length package? You either allow a situation where this variable-length encoding would not fit in a package and then it's not really clear what to do. Or you just saving bytes for nothing, because you could use fixed-length encoding instead and satellite does not care whether you sent 65 or 69 bytes of data.
At work we’ve been experimenting with what amounts to a pre-calculated/pre-shared dictionary to allow for compression of small data, but just picking what data you send and it’s representation gets you very very far.
Because we need to put nearly arbritrary data into our NB-IoT UDP socket, Inmarsat or Swarm, saving bytes let’s us pack more data into a message so we can send fewer messages: saving money and increasing reliability (because if you miss a Swarm upload time you can have hours before the next one, for example, or our Simcom modem might hit a black spot and such)
I don't think the general use case is once per minute though, more like once per hour unless something goes wrong. Something like ensure pipeline pressure is still nominal for remote parts of the pipeline.
The reason people are skeptical about its long-term longevity is that the vast majority of the income of the project is coming from the sale of hotspots, rather than from paying customers [1]. This means pricing for users is heavily subsidized by hotspot revenues right now. Eventually, the hotspot market will be saturated and prices will have to come up to keep the project afloat; it remains to be seen how competitive the pricing will be at that point.
PS: Saying LORA is "ubiquitous" in the US oversells it quite a bit. It is quite available in cities but most rural areas are completely devoid of coverage, see [2].
Oh boy, where to start with all the problems in this post.
* you can buy helium hotspots for $100 now, on par with other lora gateways
* " vast majority of the income of the project" is vague and incorrect. The hotspot OEMs (of which there are 50+) receive sales revenue, not "the project".
* You're referencing a Senet chart, which is actually a roaming partner on Helium. This isn't actually a helium coverage map.
I will correct myself on the coverage, about 98% of Americans are covered, not 98% of land in America. Oops. America is a big place.
Yeah but once a minute is overkill for lots of things - remote river flood warning every 15 minutes, river water quality once an hour, temp or rain fall once and hour
Inmarsat is surprisingly more competitive to Swarm than we expected, but man it is difficult to get passed all the bullshit resellers, which is where Swarm is way easier. But for our use cases Inmarsat has been a better fit right now, though we use both on our board depending on the deployment: if we don’t need as much bandwidth and are okay with the black out times where we’re not covered by Swarm, it’s great
Hourly data collection for science like stream gages, earthquakes, air quality, precipitation, or similar in very remote areas.
Sensors along a pipeline etc to monitor remote infrastructure.
Texting in remote areas for logging, cattle, forestry services, etc.
Redundancy for boats, emergency services etc that may have primary and secondary communication channels, but still greatly benefit from the capacity for redundant txt messages.
A wire like that is a single point of failure. Great for sending lots of data, but using two different systems is safer than using two different wires along the pipe, both of which could be severed in the same event.
It's ideal for asset tracking - you can fit a GPS fix, timestamp, and accuracy estimate (based on # of satellites visible) into 12 bytes, which is enough to ping once per minute and still only use half your allocated bandwidth. If your asset tags call in every 5 minutes or once per hour lots of options open up.
2 way communications aren't much more demanding because most of what you need can be done with lookup tables, and there aren't that many different commands you would need to send to an asset tag for local scanning. If you want to do person-to-person, 2 bytes per word is sufficient for a fairly large vocabulary (especially a tightly specified one like military brevity codes) - not exactly chatty but sufficient for many operational/emergency purposes.
Not to mention that people could have internet connected crap without even knowing it. No way to block it using PiHole, pfsense or other local routers and access points.
Given there's mobile service where most people are spending time, this situation is already much worse with cellular data b/c of the substantially higher bandwidth capability.
It's a circuit that alerts people to dead carriers.
People in radio station studios generally don't listen to the over-the-air signal because there is a delay. A silence sense is a circuit that monitors the over-the-air signal and takes action when it's been too quiet for too long. This is usually an indication that something has failed, either at the transmitter, or in the studio-transmitter link. It is sometimes triggered by dramatic pauses in classical music and talk show content, but in those cases is ignored by the DJ/host/producer.
They've been around forever, and can be made from simple analog circuits. In the stations where I've worked, if the silence sense activated, a red light lit up in the DJ booth, and the engineering department. Some stations had a secondary silence sense that would wait a bit longer, and light up a light bulb at the receptionist's desk because she had the master list of phone numbers to call the right people in case of a transmission failure.
There are thousands of radio station transmitters that are far enough away from the originating studio that it's not possible for the studio to hear the over-the-air signal, so a silence sense on top of a mountain, next to the transmitter could send an alert packet via this satellite service back to the studio to let someone know something is wrong.
This service is high latency at present and only supports sending uplink data a half-dozen times per day when satellites are overhead. It would likely not be suitable for this application.
I'm speculating, but certain FCC licenses require you to transmit without significant gaps on your licensed frequencies or be fined. (The theory being that others could be making better use of your frequency since you aren't using it.) If your transmission gear goes down for whatever reason, you want to notify the FCC before others do. E.g.,
I could see this being very useful for scientific beacons, for example to track ocean currents or bird migrations. The latter currently use cellular connections, but can only track birds through areas where reception is present.
If I was asked to engineer such a tracker, I would definitely buffer data inside to flush when connected to celluar network. Not always real time, but much more useful.
Remote communication in places without cellular connectivity. Air quality, water quality, remote weather stations, aviation, industrial applications... Latency anywhere from a few seconds to minutes. Not used like normal networking, it's something you build around for very specialized applications.
How about for check points on an ultra-marathon desert run, where you are crossing the Sahara. Each check point could have an RFID reader that gives feedback on the racers. If Bob doesn't get to the next checkpoint in say 2 hours, you know he's probably lost.
A good use of IOT is agriculture. At proper cost, having satellite connected devices across huge sprawling farms would be extremely useful. There are already solutions, like LoRA, but these have their own disadvantages.
In the normal case, you send a heartbeat once per (interval calibrated to threat model). Should be plenty of packets left to encode "send lawyers, guns and money" when necessary.
Given the low cost and power requirements of LoRa this could be fantastic wildlife and natural habitat monitoring. Could act as a backup emergency contact mechanism for folks that spend a lot of time away from civilization.
Hopefully some resellers pop up that let you buy them onesie-twosie to tinker with.
I don’t have the numbers, but at least for Iridium the data service is basically an afterthought. They allocated so much of their bandwidth to voice (using older insufficient voice codecs no less) and are now essentially locked in to a suboptimal bandwidth allocation. They must didn’t realise the importance of data when originally designing their hardware. And hardware in space is expensive to change.
Unfortunately, command and control for terrorist drone networks, sending coordinates to low-use credit-card gas pumps for night-time refills, and coordinates to targets. Give it ten years.
I love how they define "activation date" for the data plan!
> The hardware will be charged upfront but the subscription data plan will not be charged until the device is activated. The activation date (and $60/yr billing date) for a device is the first of the following month once ≥50 messages are sent. For example, if your device sends its ≥50th message on April 8th, your device would be charged on May 1st. And would renew annually on May 1st until you opt-out.
I have one of their dev kits (the neat one with a built in solar panel and Wi-Fi connection) - sadly I was unable to get it to send/receive more than 1 message in San Francisco over a 3 month period.
(I live on a side of a hill and had the entire dev kit standing on my roof that was tall enough to have a view over a large part of the city and out into the bay)
At the time, I heard from support that they were launching new satellites that would improve the connectivity in built up locations (such as a city) but I eventually gave up.
At the time, there were 2 satellites a day that were prospects to connect to based on my location, but we met an invisible issue that couldn't be debugged. Potentially it was a satellite visibility issue but could have also been that the available satellite connection was backlogged with other devices / the satellite didn't have a large enough window of connectivity to perform the upload / there was noise interfering with the connection / other mysterious reason that would need expensive test equipment to root cause.
I took it with me into the plains one day (Livermore) and it connected once in the 8.5 hours I was there so it was functional and the RF signal strength was good for the received packet.
My takeaway was that the tech is well suited for rural areas all over the globe where there is no other connectivity options and there exists applications that can work well with the high latency / low bandwidth payloads. Well monitoring jumps to mind as does CO2 gas monitoring in mines - slow moving signals, rural locations, expensive to send a truck to go and monitor.
Have you tried panel mounting the antenna on a big (at least 4 sq ft) ground plane? We use these at work and had a very hard time until we started building nice big ground planes into all mechanical designs (that and reading the design advice on the decoupling network with a microscope).
Huh, for some reason, I thought that the FCC would fine and regulate this company into oblivion after their unauthorized launch stunt. Looks like that didn't quite happen the way I expected.
Same, it's weird that this hasn't already been mentioned. They launched untrackable trash into space and got fined for it, nobody should trust this company.
It drives me up a wall that they got off basically consequence-free for that--I'm guess they probably made up the fine in reduced time cost spent on authorization proceedings anyway. "Move fast and break things" isn't a great motto anyway, but it's especially bad when those things are floating around in orbit.
- IMHO, price could be prohibitive for many use cases (e.g. water meters)
- Signal strength could be an issue for urban environments (e.g. inside buildings)
- There aren't many small business use cases for IoT connectivity that does beyond WiFi (both wrt range and power budget). What I have seen is that there are usually several big players (e.g. utilities, municipality, factories) that at some point deploy LoRa based or similar connectivity. LoRa can actually cover a huge area with just a handful of self operated gateways (similar to Helium in this respect).
This reminds me of Hiber[1]. Hiber launched in 2016 with the mission of connecting the 90% of the world that didn’t have access to connectivity. They were working on HiberBand, a protocol to communicate ultra-low power for remote IoT solutions. It sounded very promising and futuristic, a bit like a mix between Starlink and Swarm of today.
At some point Hiber pivoted away from a connectivity-only business model. You could no longer use them for general purpose IoT connectivity, instead they focused on building end-to-end solutions. Hiber today markets itself as an "off-grid asset monitoring" solution focusing heavily on the oil industry.
I suspect there is some underlying issue with these high-latency / low-bandwidth solutions that just makes them impractical in real life.
Ultimately what this means is that data transfer speeds will be faster, yes? (That's a question.) If most internet usage is direct to satellite and then satellite to endpoint this would mean that tier one hub networks would no longer be used. So laser replaces fiber for most of the distance, with a peer to peer satellite network that would make outages less common. Which would be both faster and more reliable. The only issue I see is for there to be an intentional or unintentional satellite debris event that would be self propagating, but that issue has been around for a while anyway.
What exactly is the plan here for when a satellite is obsolete or dies?
Currently we either have the satellite re-enter earths atmosphere or we use the last bit of fuel to send them even fruther away. But these seems very static
If these truly are just 11x11x2.8cm - then in a few years we will have bunch of super fast and very destructive mini bullets orbiting earth and destroying very expensive space equipment.
Can we please consider the amount of trash we send out in space before it is too late?
Pretty cool. Just checked coverage over my house using their pass checker and it's already ~50% coverage through the course of the day. No good for realtime alerting but good enough for daily reporting.
Yep, it seems like you get a transmit window every ~2 hours max. All in all this service is much more reasonable(cost/antenna size/transmit rate/power) than I expected. You can basically send 2kb/day from a large chunk of the world(with a view of the sky) for $5/mo. Somehow even cheaper than m2m cell plans I've encountered, though granted I haven't looked at services in the past few years.
What does the data collection end of this look like? In a home IOT setup, for example, you'd typically have a single machine (Border Router in a Thread setup) collecting all the data from each node.
I understand the "field device" -> "satellite" uplink step. But, what's the next step? Another device just for downlink data would eat through the data cap in no time.
Looks like they have a SaaS product called Hive to poll/push data from devices to your application. I don't see any indication of point to point message passing capability but it may exist.
> Swarm uses a narrow band carrier (about 40 kHz wide), and a chirped spread spectrum approach
Is this LoRa? I guess if it was they wouldn't need their own modems. There have been some hobbyist attempts to put LoRa gateways on satellites and from what I remember it worked reasonably well.
High. At least if it's anything like Iridium, probably seconds to minutes depending on coverage. Really not designed to be used like general purpose networking.
Well this is terrifying. Pretty soon you'll have to crack open your appliances to snip the antennas to prevent them from calling home, rather than just not giving them the wifi password...!
You're already carrying around a device that monitors your location, what you are saying (and audio in the surrounding area), to some extent how you are moving (gyroscope), your data ingress and egress patterns and where you share virtually all your public thoughts (and most private thoughts).
Some people even connect their watches, which monitor (or will soon monitor) everything from their heart rate, sleep schedule, oxygen levels, BMI, etc.
I frankly don't think it can get much more intrusive.
This strikes me as defeatist. You can run GrapheneOS and pipe everything through VPNs for your phone. It's really easy, the phone does all the normal stuff except the constant listening ("assistant") thing, and it's way harder for anyone interested to spy on me. I don't run Google or Apple software.
If someone could get their hands on my data, it'd be a bunch of extra work to determine who it's coming from.
>>Some people even connect their watches, which monitor (or will soon monitor) everything from their heart rate, sleep schedule, oxygen levels, BMI, etc.
You can easily have a watch that does all these things but doesn't share any data, and you can be sure of that based on more than just promises. Open hardware, open software, local data storage and zero-knowledge for anything touching the internet. It's well within our grasp technically, we the people just need to demand it.
LoRa's much lower energy, maybe? I can see cell networks if it's plugged in. Or maybe cell radios are cheap enough that you could put one in and fire it up once a day to squirt out data on how to better sell your kids sugary cereals...
This is already possible with cell connectivity. Many cars have this now, for instance. Nothing is stopping the telcos from offering something to compete with this other than in remote tower-free locations.
And heavy equipment had this available years ago. One-time fee up front and good for the life of the machine. Can be installed in any type of machine. Worldwide GSM and can send commands to the machine. Was around USD$1500, IIRC.
https://www.komatsu.eu/en/komtrax
I literally watched a large excavator from a dealer break sanctions and show up digging in Iran.
Surprised dedicated ultra low bandwidth satellites are the way to go when developing a service like this versus piggybacking on something like StarLink.
This was a separate company originally. In this case the RF protocol is completely different (~140MHz vs 12+GHz) as well so you'd have to add hardware to the StarLink satellite to support this (which obv. could be done moving forward)
No? SpaceX purchased it, so it's theirs now. They may for instance be using shared ground network hardware and software post acquisition. Do you have reason to think otherwise?
this is actually really great for applications like tracking, remote instrumentation, and other services that have a small amount of relatively important data in hard-to-service or critical applications which is not that wildly specific.
Lots of sensors with a restricted pipe isn't uncommon - in the electrical transmission substation world they've been dealing with the problem of how to do a lot with a tiny amount of bandwidth for decades. An industrial fieldbus protocol even exists for it called DNP3.
The answer is your end devices only signal on meaningful change, whatever that may be, and stay quiet the rest of the time. How meaningful that change must be depends on how much bandwidth you have. When you get right down to it you can do a lot with a little in the industrial controls and remote monitoring space.
We get so used to overly verbose communications protocols and overly abstracted data structures and methods, it's refreshing once in a while to care about each bit!
Your hardware can record many datapoints, and bundle them up together into one packet for transmission in one big bundle. If you are reading a float off a sensor, you can get nearly 1 data point per minute, transmitted hourly, with a very naive implementation.
Much more with compression specialized to your use case and clever packet design.
What do you mean 1 data point? It transmits bytes, not data points. I could easily stuff 60 temperature and humidity readings into one hourly 192b packet(ie per-minute weather data). That's 120 data points, not 1. And you could easily double that with simple differential encoding or beyond other compression techniques
Yeah what I meant to underline was that for a critical application you won't be able to react faster than one hour in case the data reports a critical issue. So not suitable for security or any application where you need less than one hour latency. I guess if you need satellite coms maybe the sensor is not in a place you can go in one hour anyway.
You aren't going to be streaming video from the Atacama through this but it is no where near useless. 192 bytes is actually substantial amount of data depending on what you're interested in.
ADS-B is essentially mandatory nowadays, but it's just a radio that transmits location to others in the same local area. Get enough volunteers with receivers spread throughout the world and you do get real time location tracking of every aircraft. See https://www.adsbexchange.com/
Can already be done for a few orders of magnitude cheaper (and with several orders of magnitude more bandwidth) with a cellular connection. The people who want to track you/your data aren't going to use satellites, they're going to use Verizon or T-Mobile or AT&T.
