WiFi and Bluetooth were never meant to be the radios used by a billion Internet of Things hats, umbrellas, irrigation systems, or any other device that makes a worldwide network of things interesting. The best radio for IoT is something lightweight which operates in the sub-Gigahertz range, doesn’t need a lot of bandwidth, and doesn’t suck down the power like WiFi. For the last few years, a new low-power wireless communication standard has been coming on the scene, and now this protocol — LoRa — will soon be available in an Arduino form factor.
The Primo, and NRF
It’s not LoRa, but the Arduino Primo line is based on the ESP8266 WiFi chip and a Nordic nRF52832 for Bluetooth. The Primo comes in the ever-familiar Arduino form factor, but it isn’t meant to be an ‘Internet of Things’ device. Instead, it’s a microcontroller for devices that need to be on the Internet.
Also on display at CES this year is the Primo Core which we first saw at BAMF back in May. It’s a board barely larger than a US quarter that has a few tricks up its sleeve. The Primo Core is built around the nRF52832, and adds humidity, temperature, 3-axis magnetometer and a 3-axis accelerometer to a square inch of fiberglass.
The Primo Core has a few mechanical tricks up its sleeve. Those castellated pins around the circumference can be soldered to the Alice Pad, a breakout board that adds a USB port and LiPo battery charger.
LoRa
Also on deck at the Arduino suite were two LoRa shields. In collobration with Semtech, Arduino will be releasing the pair of LoRa shields later this year. The first, the Node Shield, is about as simple as it can get — it’s simply a shield with a LoRa radio and a few connectors. The second, the Gateway Shield, does what it says on the tin: it’s designed to be a gateway from other Arduino devices (Ethernet or WiFi, for example) to a Node shield. The boards weren’t completely populated, but from what I could see, the Gateway shield is significantly more capable with support for a GPS chipset and antenna.
A partnership with Cayenne and MyDevices
Of course, the Internet of Things is worthless if you can’t manage it easily. Arduino has struck up a partnership with MyDevices to turn a bunch of low-bandwidth radio and serial connections into something easy to use. Already, we’ve seen a few builds and projects using MyDevices, but the demos I was shown were extremely easy to understand, even if there were far too many devices in the room.
All of this is great news if you’re working on the next great Internet of Things thing. The Primo Core is one of the smallest wireless microcontroller devices I’ve seen, and the addition of LoRa Arduino shields means we may actually see useful low-bandwidth networks in the very near future.
LoRa sucks because it’s not open. Will they ever learn that? ZigBee has “failed” while nRF, BLE and ESP modules rock the world.
Zigbee failed?
every single non hacky smart home system is Zigbee based. ZHA utterly owns the home automation market. Crestron, AMX,C4,SmartThings,Wink,Hue,GE,etc… all of it is a zigbee running the super closed ZHA protocol.
If that is a failure, where do I sign up to be that kind of failure?
Sure, they have market share, somewhere.
But they aren’t even considered by the current graduates. In tkme, they will fail.
Closed source money grabs only work for a short period of time.
You might be a little out of touch.
When I graduated from FAU about two years ago, engineering student projects were using zigbee for pretty much everything, from senior design projects to graduate research projects. My senior design team ended up ditching wifi for zigbee because of logistical constraints.
Maybe FAU has someone in management pushing the things for one reason or another eh.
“Closed source money grabs only work for a short period of time.”
Really, how long have Windows and Office and MacOS and iOS been working ?
Zigbee fails because it is almost the most insecure thing you can use. Market opportunity?
It uses an exotic frequency (~900Mhz) and supports encryption and it’s easy to know on the box what actually uses encryption or not and it NEVER used pump to open your network to the entire world. Might not be perfect but it’s orders of magnitude better tjen wifi or Bluetooth trash products. Unless you are streaming audio or video anything other then zwave/ZigBee is trash.
Could you please spell and grammar check your comments before you post them? You’re pretty much incomprehensible because of that and your writing style.
I’ve only and will be buying more zwave/ZigBee devices simply for the fact they don’t fight for 2.4 spectrum in a world super saturated in wifi I’d rather my lock not waste more battery power trying to be the loudest device.
Um, ZigBee is a layer on top of 802.15.4 – which is USUALLY (with rare exceptions) in the 2.4 GHz cesspool.
That said, like nRF 2.4 solutions and TI 2.4 solutions, it’s low spectral bandwidth stuff that performs very well in “typical” 2.4 interference scenarios.
TI’s proprietary 2.4 solutions utterly destroy BLE for any IoT/sensor network solution for example.
Call me impulsive, but I’m betting LoRa will do pretty damn well.
Microchip are investing hugely in it. SigFox have got a good thing going on. Not to mention all the other investment being thrown at it.
I think it was a 33C3 talk from this year that had some slides showing just how much money is on the table already here.
LoRa might be closed, but it already being deployed all over the place, to great success.
Also, that talk was paving the way to opening that closed nature.
https://media.ccc.de/v/33c3-7945-decoding_the_lora_phy
It’s a trash protocol why on earth would I want my home automation competing with my laptop’s and phones? Use another frequency!!! (Zwave is on 900 range)
Trash protocol? Care to back that up with data, or should we just take your word for it? The whitepapers I’ve seen show incredible results for low data-rates over long distances at low power. And for the record, LoRa is not tied to any frequency (i.e. 2.4 GHz), I’ve seen modules for 400, 800, 900 MHz.
I’m not biting. Do some research, then come back to the discussion.
Know what you’re arguing, compare like-for-like.
What has protocol to do with frequencies?
And, BTW, LoRa uses ISM which, combined with adaptive data rate and channel hopping, is more information-density efficient and has better chances to grab airtime than “classic” UNB methods.
LoRa is not perfect but the spec is open and there is considerable momentum from hardware vendors and open back-end projects (remember, the sensor is NOT the most important part in an LPWAN: it’s the back-end).
This is the dumbest thing I’ve heard in a long time.
Hey, I want to support open standards, but when you’re open, you have to MANAGE that openness to ensure interoperability in some way. We’re lucky that even without any significant enforcement authority, the IETF has enough reputation to ensure that someone who puts out noncompliant shit is going to self-destruct their own reputation. But IETF standards/RFCs are the exception and not the norm in this regard.
Yes, ZigBee has pretty much failed outside of ZLL – and that’s not due to lack of openness. ZigBee is far more open than any of the other HA standards (yes, it has licensing issues, but at least the spec is public), and in some ways – that has been its problem. They haven’t done any significant amount of interoperability enforcement and as a result, it’s a gigantic mess of poor interoperability that can be partly traced directly to its openness. (See TCP/Greenwave ZLL units, for example). Note that no ZigBee HA implementation is marketed as primarily as being ZigBee – it’s marketed as being Hue compatible. That’s because ZLL certification is meaningless (again, see TCP/Greenwave’s certified but noninteroperable crap collecting dust in a closet of mine somewhere…) with regards to interoperability.
There’s nothing else using ZigBee out there – well maybe some of the Lowes Iris products are ZigBee – but the Iris products that use ZigBee are only marketed as Iris-compatible. Contrast to the Iris products that conform to other standards with greater interoperability track records.
What dominates HA? Z-wave right now, with maybe Insteon being next. Unlike the ZigBee crap, Iris Z-Wave products are clearly marked as being Z-wave compatible. Both Z-Wave and Insteon are VERY closed, far moreso than ZigBee. But they dominate their industries.
nRF? Wait – you’re saying that a mostly-open standard will be supplanted by a single-vendor 100% manufacturer-proprietary solution? Just because anyone can buy Nordic’s stuff off the shelf dirt-cheap doesn’t mean it isn’t also 100% proprietary. Same with some of TI’s CC-family stuff (such as that used by Dexcom’s G4 CGM system).
BLE? Not well suited for IoT/sensor networks/anything that is “mesh-ish”. Also sucks rancid infected donkeyballs for anything extremely lowpower such as many sensor network applications. See Dexcom’s G4 system based on proprietary TI RF hardware vs. their G5 mobile system based on BLE. Twice the physical size (all due to battery) half the rated battery life (3 months vs 6), and around 1/4 the typical actual battery life (many people including myself have had G4 transmitters last over a full year) – all with vastly inferior range and reliability despite having a protocol that includes ACKs and retransmissions.
ESP? You just indicated Yet Another Vendor-Specific Product – yes at least it’s 802.11, but that makes it poorly suited for many IoT tasks due to the limitations of 802.11 for such tasks. (Oh yeah, from what I’ve seen, the ESP8266 sucks for mesh applications due to limitations regarding inability to forward traffic between multiple clients associated to its AP when in AP mode.)
Faster Arduinos, where are they?
Are my next Arduinos are going to be Teensy 3.5s?
Try the NodeMCU – works great with Arduino IDE and is pretty fast at 160MHz.
I heard on the down-low from one of the Atmel guys that there was a Cortex-m4 design in the works, rumoured to be an Atmel part. Shows how long ago I heard this!
Not sure what’s happening now MCP are running the show. Arduino & Atmel were really friendly, with a lot of support given to the maker community, even if it was through back channels.
Yes, there is the MicrochipMakes mokier being thrown about, but… Well. I’m just going to sit back and watch this pan out.
I’d bet a few dollars on the next few boards being Nordic based too.
there is an STM32 ARM CM4 in the works, which I am excited about because Atmel’s ARM line isn’t very good compared to STM32 NXP LPC or even FreeScale Kinetis. The Atmel ARM line is overpriced and the controllers are not as peripheral or feature rich as their competitors. I really hope they ditch Atmel entirely for the ARM versions and go STM32 its a superior platform and has way more developmental tool support both open source and closed.
Cool. Now, Arduino, start working on adapters that allow those round thingies to be plugged into arduino shields. And make sure the new round pins don’t accidentally end up usinh any standard offsets.
Is that a bodge wire I spy on the full-size Arduino Primo?
It is, it is!
Two in fact!
(Apologies for double post. edit button anyone?)
MM. I’m taking a punt, based on the package size, and guessing that it’s the IMU is a punt itself, but is that a BNO055 IC, from Bosch? Or one of that family?
Fingers crossed, it’s a tasty device.
Can anyone tell me why it would be a good idea to have a round board with castellations such as the PRIMO CORE? I guess one can say watch applications or circular tags but the thickness, arrangement, and board components don’t mesh quite well with the goal of a small design that could be better accomplished with a rectangular form-factor broken out onto a circular PCB.
It’s an attempt to mimic / cash in on the craze started by lilypad for round uC boards (which was done to make it more easily integrated into clothing/wearables.)
That is such a fail, what on earth does a round board with castellations accomplish? You can’t sew conductive thread into it, there is no hole to secure/tie it. The footprint will be a small nightmare to make pads for, to solder it onto PCB’s. All I can figure is they want you to solder wires to the castellations, otherwise that layout has little headroom moving forward into future designs.
Talk to me when your using a logical power efficient protocol. (Zwave+/ZigBee)
Unlike wifi products your have a level of quality control and you can read the package to know it it supports encryption or not. My kwikset lock isn’t perfect but its better then ANY wifi or Bluetooth option my battery’s last over a year on a single charge and it doesn’t compete with my wifi like crappy wifi home automation does. Unless your dumb enough to not segregate by our network. Best part about zwave is no upnp.
Bluetooth LE WAS designed for IoT. Your statement that Bluetooth was not designed for IoT is not quite correct. Yes classic Bluetooth was not, but LE WAS!
https://www.bluetooth.com/what-is-bluetooth-technology/how-it-works/low-energy
While any attempts to educate are indeed noble, may I suggest not feeding the trolls ;)
It’s pretty clear from these ramblings that this commenter is not well versed in what they are talking about, and the steadfast attitude exhibited shows little inclination to change that fact, unfortunately.
Everyone who is shocked that they are using LoRa, remember when Microchip bought Atmel? Well they make a bunch of LoRa hardware, would not be surprised if they encouraged Arduino to use it.
What is the round board in the first picture? It says NFC on the backside but I can find no reference to it on the web or on the Arduino website. Article does not seem to be connected to the pictures.