I’m a regular user of different PCB services form my electronic design. Since 4 year now I’m mostly working with Eurocircuit but recently I tested different low cost solutions : seeed studio, jlcpcb and OSH Park. This is now an opportunity to compare these different solutions.
New episode of my LPWAn post series, this time, about the network architecture and the role of the network kernel.
The network architecture is different than a classical IP communication from client to server: the LPWAn architecture is based on a kernel network allowing the protocol transformation from the Sigfox / LoRaWan world to the IP-Internet world.
In the LPWAn protocol the addressing is limited to one device address only to reduce the frame size and as a consequence all the frame have to be captured by a central system: the network kernel. It will route the message correctly to the end user application based on device association made on registration.
This post is related to a french video (as usual) and detailed in the following part of this post.
TD1208/TD1508
TD1508 is the FCC version of the TD1208, they are pin to pin compatible and based on the same SDK so converting an object working on a TD1208 sounds easy. Practically speaking there are some preparing steps because of some tiny bugs to solve.
Let see how to port a Telecom Design existing firmware for TD1208 to TD1508 (from ETSI to FCC).
The Things Network (TTN) is a global LoRaWan public network kernel based on crowd-source infrastructure. This initiative sounds really interesting in my point of view because it breaks one of the main LoRaWan issues : the country based operators organization.
Related to this good point, the negative point is the current area covered by a such network limited to the contributor localization, generally in the main cities.
That said, I’m convinced : when you want to create a private LoRaWan network you have to manage a network kernel managing gateways, device authorizations… You can do it yourself, you can pay someone to do it for you or, eventually, you can lever TTN for this use. As a counterpart you will extend the network and offer this benefit to any around. That way it makes sense and gives large opportunities to the networks and it’s private users.
Long introduction for a technical topic … How to join this network once you have a Kerlink LoRa Iot station available ? Continue reading
Sigfox has announced end of last week the opening of the device Geo-location service. This solution allows to get the location of a device based on its radio signal.
Different announced has been made in this domain by Sigfox and its competitors LoRaWan. The technics in place are not the same and the Sigfox solution is based on BigData.
From the beginning Sigfox was returning a device location based on the signal measured by the different base-stations. This location was previously rounded to the nearest lat/lng degree. This information has a precision of 100km x 150km basically. Even if it was not possible to locate a device precisely, i was interesting to know in what country corner it was located.
With the service launched in February 2017 Sigfox is moving to a more precise location based on radio signal, according to different discussions I had the precision goes from a couple of meters to some kilometers depending on situation, data already gathered at this position, number of antennas …
Business model is in my point of view a big blocker for IoT tsunami : when IoT is service and service is monthly fee, IoT is also objects and objects business model is based on push the box and get 30% standard business model.
These two approaches are really different and distribution must change its business model to correctly fit with IoT and extract the right level of value. Taking 30% one shot is quick & easy but getting a regular revenue based on service delivery is more profitable.
The blocking point seem to simply be process, people and software: actually distribution is not organized and equipped with the right IT for managing recurrent revenue, for this reason the distribution have no major interest is service oriented product and prefer all inclusive device where service is included in the initial price.
The problem of this model is that it is not profitable for IoT industry : IoT have a making cost to produce the object but year after year each object generate a service cost. In most of the existing business case, new object sell are funding the service for old objects. This can work when the company is growing and make scale effect reduction on service cost. The problem start as soon as the market is decreasing and the cost of running service is higher than the margin made on new object sell. This problem is a killer as soon as your service cost is high like when you have to support a recurrent communication fee (like for what I consider real connected object, I mean objects able to communicate with no local infrastructure dependencies). The only way is to make money on data collected from the object and be payed on client privacy. This equation is really dangerous for the client, even for the IoT industry as much as privacy regulation law are growing and data market not really predictable.
I’m quite sure that this situation will soon change as distribution can make much more money based on service than based on box pushing and as soon as they will be organized to manage it, as the business is better, they will make a lot of available space for service business model based object, enabling the IoT tsunami.
The downlink process is the way a sensor can receive data from the network. This process is really useful if you need the device to execute an action or if you want to change the sensor behavior.
This post is as some other of my LpWan series a video post in French completed by this text post in english.
The Snoc Wisol DevKit
Basically SNOC SFM10R1 aka BRKWS01 Sigfox devkit is an easy way to get start on low cost sigfox hardware thanks to the low cost module from wisol based on the low cost chip from On Semi Ax8052. Presented like this it looks like stack on a stack on a stack.
But in my point of view the solution is really more than this. As a maker I know how difficult it is to kickstart a Sigfox project : you have to deal with expensive devkit where you sometime need a Kbis, then you have to deal with sigfox to get an access and at ends if you want to transform you POC in a product you need to change you technologies for cheaper or simper solutions. (here I’m not pointing anyone but basically I had to made my own devkit to start on sigfox) Most of these problems have already been solved thanks to sigfox making account creation so simple for devkit and the arrival of devkit like snootlab, airboard or smart-everything. All these solutions are nice but provides a full stack solution MCU + communication for a price around $100. This is where the SNOC module is proposing an alternative : get an easy access to the sigfox network with a low cost solution you can plug to the MCU of your choice… Arduino, Rpi, Nucleo, Pic it’s up to you !