WiFithing is making an IoT development platform with a twist, low power radios and a back-end. While SemiAccurate doesn’t normally talk about Kickstarter projects, this one seems to have merit.
Most IoT platforms are from a big vendor with tools to make software and hardware development simple but usually only for the products they make. WiFithing is a layer higher than that, the tools, platform, and web-based back-end are already there. If their offerings are what they claim, and they look to be, things should be a lot easier and possible even to the level of their “Code Free IoT” promise. Luckily for almost everyone involved the offerings provided are all open source so if you do want to code, you are not trapped in a proprietary walled garden.
The hardware is the easy part to describe, it consists of two types of boxes, masters and slaves. As you might have guessed the master is effectively a gateway, the slave an endpoint device. Unlike what their name suggests, the main communications protocol for WiFithing is not actually Wi-Fi but a TI ISM radio link instead. That said it will use Wi-Fi to connect to any Wi-Fi device out there, and with the right drivers can control them like a WiFithing slaves too, but the main idea is to use a slave over ISM to save power.
As you probably have heard by now IoT devices tend to be very low-bandwidth and bursty in their communications. Wi-Fi overhead is pretty painful in such cases and even the newer protocols with low energy in their names are only lower energy than Wi-Fi, bespoke protocols can do much better. That is the point of WiFithing, lower energy comms and open hardware.
The boards without the white plastic box
If you look at the master board it has a lot of pinouts, 16 and SPI on the master, 23 on the slave. These can be used for almost anything, anyone familiar with Arduino/Energia will be at home here. Both units can be powered by batteries or DC/USB, battery life depends on what you use it for but should be pretty good regardless. Each master can control eight slaves or a larger number of bespoke devices. If you want to do something different, you can flash both ends, the pins are open and the code is available.
Speaking of code that is where WiFithing steps out from most development platforms, it is effectively ready for the end-user rather than an enthusiast or development platform for coders. You can see this with their authentication mechanism, a QR code on the devices. You scan it with your phone and it brings you their web back-end. This allows you to pair a device with a master that you presumably own, then pushes the code to the device via the master’s net connection. If you don’t want to manually flash a device, you don’t have to.
From there on out any changes to any of the devices can be made from the net, you don’t have to do much if anything to the master or slave. You can assign names and functions to devices via a drop down menu, and they will then appear as services you can select from the main screen. If you put in a thermostat and tag it as such, it will then appear as one on the main screen and offer you the appropriate functions. If there isn’t a function for you, well you can write one if you need to. You can see a demo of all of this here.
With just a little soldering WiFithings should allow a user to do almost anything they need. The code is mostly if not all done, the pins are open and exposed, and so is the code. If you want to use the WiFithing back-end, they are going to charge you $18/year per location regardless of the device count. That isn’t bad, $1.50 a month is not a large ask by any means but if you don’t want to pay, you can run your own service, the code is open.
At the moment WiFithing is in the midst of their Kickstarter campaign. As of this writing they were almost half way to their extremely modest goal of $22,081. A master costs $57, a slave far less, and at the moment a $56 pledge on Kickstarter will get you both and six months of the back-end. For the big spenders you can get two masters, eight slaves, a year of back-end service, and a meet and greet with the team for only $315, not a bad deal at all. We hope they meet their goals, it looks like a solid IoT platform for the mildly advanced user, not the hardcore developer.S|A
Latest posts by Charlie Demerjian (see all)
- More on Intel’s 10nm process problems - Sep 17, 2018
- Intel puts out another 14nm 2020 server platform - Sep 11, 2018
- Why Can’t Intel Supply Enough 14nm Xeons? - Sep 10, 2018
- Intel can’t supply 14nm Xeons, HPE directly recommends AMD Epyc - Sep 7, 2018
- AMD reintroduces the Athlon name with two CPUs - Sep 6, 2018