I have a project I’m working on that needs GPS, BLE, and cellular connectivity. I have never designed with wireless devices before, so I am wondering if anyone has any suggestions for easy to use and implement solutions that will get me GPS, BLE, and cellular connectivity?
Ideally it would be all in one module but if using multiple different modules is easier than that’s fine. I just need something simple and straightforward to design in.
Any and all thoughts, questions, or ideas are welcome… this is uncharted territory for me!
We use Quectel modules, many integrate the cellular and GNSS, not certain is there one that includes BLE, the technical support is great (at least here in Mexico).
I also use Quectel and agree the support is great. They will help with layout and antenna reviews. They do have Wifi/BT modules, but I’m not sure they have any integrated with cellular. All the cellular devices have GNSS.
One of my clients is Blues Wireless. They have a GPS+Cellular offering that I think you might want to look at. The stuff I’ve built for them pairs their module with an ESP-32 which provides BLE and WiFi.
Their value proposition is that you just buy their device and it comes pre-provisioned with network access, and it provides a simple interface to handle the communication/data transfer. It’s aimed for low-bandwidth M2M applications.
There’s many options for this. If for instance you will be selling outside the US, it’s a major hassle to handle roaming for devices unless you sell many thousand units. Due to this, companies such as https://particle.io and https://blues.io offer M2 modules that are easy to integrate into products. It’s all a question of convenience vs effort and for most smaller vendors, it’s just a lot of hassle.
And don’t forget - “Cellular” no longer means a data module with a SIM. You could use 4G, LTE-M or NB-IoT depending on your use case. In general - if your device is primarily stationary & battery operated and you do not need high data rates, I’d look at NB-IoT. Otherwise, LTE-M is liekly a better alternative. Both offers better coverage than 4G/5G data modules.
The Blues Wireless seems like a great option.
Chris G just published a TheAmphour interview (#603) with the founder of Blues Wireless, Ray Ozzie.
I myself have worked with Nordic Semiconductors nr9160 for modern cellular connection (LTE-M/NBIoT) and with different ESP32 modules, most recently with the S2 (just wifi no bluetooth) the C3 and S3 (the latter two with wifi and bluetooth). For RF I favor pre-certified modules like the esp32 modules (not just a ESp32 chip but a SMT module with ESp32, supporting parts plus an antenna or antenna connector) Using pre-certified modules could significantly reduce certification costs of your finished product. Nordic Semiconductor’s classic nrf5 series chips are of course another popular choice for Bluetooth-centric solutions.
A common challenge with cellular applications is to find a good operator, that can provide good affordable data coverage in the regions or countries your device . There are often issues with onboarding of users and fleet management of sim cards (e-SIMs is no different).
I have no personal experience with Blues Wireless, but I was impressed with what I learned from Chris interview with Ozzie. Things like that it works in 130+(?) countries, that the cost for a large chunk data and cloud service is bundled in a relatively low per unit price. If all that works as advertised, my only remaining reservations would be around onboarding of users and fleet management. Another issue might be the M2 form factor and connectors, that might not fit all applications.
There are several “assisted GPS” offerings these days where wireless modules also offers some “assisted” GPS capability like the cellular nr9160 from Nordic and the recent Lora-chips in the Semtech Lr11xx series. But all “assisted” GPS solutions use some cloud service that you need to connect to and exchange data with - at a subscription cost and data usage. These devices can not do standalone GPS, it has to offload GPS satellite calculations and different kinds of atmospheric compensations to a cloud service (another approach is differential GPS). U-Blox is player with a mature lineup of standalone and collaborating GPS solutions. I have used U-Blox extensively for things like animal trackers in the past (think FitBit+LoJack for horses). I even built GPS testing equipment for lab and production, a closed RF-box with an DIY GPS simulator for about a thousand dollars per box - based on SDRs. We needed 4 test rigs. The test equipment that big companies bought from e g Rohde Schwarz started at 50 thousand USD. But you have to build proper RF-safe test boxes, since transmitting false GPS data in open air (our testdata in the GPS satellites simulator) is a serious offense (you risk imprisonment) in most countries.
Thank you for you help… yes, I think NB-IoT is what I am looking for.
Thanks! I will check them out
If you know Zephyr very well (see note at bottom), the Nordic 9160 based modules, of which there are many, might work for you if either your GNSS or LTE requirements are intermittent. If so, these are the best choice.
If you have concurrent GNSS and LTE requirements, uBlox has the best solution. They are also easier to program than systems with Zephyr unless you have extensive experience in Linux devops.
If you’re an embedded dev, adding an nrf52 to an uBlox module will be easier than adding a uBlox GNSS to an nRF9160. The reason is: Zephyr. Don’t buy the hype. I’ve deployed consumer products with Zephyr, and I would still choose not to use it on a new project if what I’m doing isn’t cookie-cutter. Unfortunately, 9160 can’t do concurrent gnss and lte, so you’re venturing away from cookie-cutter by default.
Will this device have low power (battery) requirements?
How good does location have to be?
I was so impressed by the blues.io product, which I learned about for the first time on The Amp Hour interview, that I ordered a demo unit immediately. This despite South Korea not being an officially supported country (there are some workarounds) and that I don’t have a clear use case.
I never would have imagined that providing remote device connectivity could be this inexpensive and painless. I do have a couple of clients who might benefit, but nothing at volume. I’m really curious to experiment with my module (and supplemental SIM cards) when they arrive later this week. Alas, @ToyBuilder, I ordered the Raspberry Pi dev kit, not the one you are working on (I had a shaky introduction to using the ESPxxxx family of chips, and I subconsciously shy away from them when I can).
I’m biased, but there are Particle modules on the horizon that will do all you need, and can be integrated easily. (Tracker-M).
Particle and Blues are direct competitors so I won’t sell it too much here, but I’d be happy to go into more details if you are interested.
(I’m the team lead for the TAM team at Particle)
+1 for Particle modules. They’re a great company and have the full stack support for the modules. Recommended for shaving months off your design.
If you feel like going it alone, the ESP32 is a great BLE module.
I personally prefer U-blox modules. They have great price breaks and support.
Yeah. Particle used to be my #1 goto, but I don’t like how they’ve treated my customers (I design/prototype, they sell). I wrote a bit about why I fell out of love with Particle on my blog, but it boils down to prioritizing new customers over existing. Our devices kept breaking since the Particle SDK/firmware wasn’t tested well enough against the P1 moules that we now have distributed across the world. For an IoT technology provider like Particle, this really is a no-no and the recent price hikes was the final nail for the two big customers I had on Particle. You just cannot break existing customers to chase new tech & new customers…
That’s why I now use other alternatives. If I just needed cellular connectivity, I’d look at Blues for now. For Wifi, I’d want to make my own hardware and backend. Particle’s backend & pricing does not scale well for high bandwidth solutions such as realtime reporting & transactions like my customers mostly do. The Particle backend had 3-5 seconds latency some days. It’s just not realtime like you can get with AWS and given that we pay for the Particle backend and only really use the OTA, we might as well roll our own based on what Espressif and lots of others supply.
I personally don’t like the scaling options with Particle. When we wanted to scale our product, Particle wanted a minimum business guarantee. They wanted us to enter into an agreement that simply wasn’t practical.
Our business just wants to order ‘n’ units and start installing it. They don’t want to deal with the hassle from both sides.
While Particle had the best infrastructure to build a product around, their business models are generally very difficult to work with.