In my definition an IoT device should have its own IP address and be directly accesible through the net.
Looks like you use bluetooth as wireless communication channel so the devices are not IoT devices themselves.
So you always need a bridge to use the devices?
The idea is very nice, ofcourse a lot has been built already in the hobby community, it is very nice that you made the SW so easy to use that kids can use it as Lego blocks to create and it would be a nice addition to the Lego assortment esp. if prices can be kept low.
What I do miss is a nice plastick box around the devices now they are a bit too naked and vulnerable for kids. Good luck on your project.
I do agree with the fact that a true IoT device should be exactly that, directly accessible from it's own IP, IPv6 perhaps, as Dave points out. However, for this particular project wi-fi is not a good fit, both in terms of price and specs. It's far too expensive and also power hungry for a 'real time' sensor; these are not blocks that are just accessible from the internet, they must be accessible locally as well.
In my opinion, unless wi-fi addresses the cost issue, which slowly it kind of trying to do, BLE will continue to expand. We might even see a home internet router which embeds BLE as well, that involves some sort of direct BLE addressing over the internet?!.
The PCBs will be conformal coated to avoid potential shorts, water damage and so on but if we get enough money and have enough time we will probably design a simple casing. We have a couple of casing designs ready but nothing too 'realistic' at the moment. Thanks for your feedback.
Since no schematics are available yet, I only have a couple questions on the implementation based on what I learned at Maker Faire. First, why did you go with a lipo + usb charger solution? Seems like an expensive option. Did a coin cell not provide enough battery life, or was it instantaneous current requirement or some other decision point. What is the battery life? I know you are using a CSR part, can you tell us if this is using the CSRMesh software, or is it strictly point to point?
We didn't really want to put the user in the position of having to buy batteries all the time. It might be cheaper for us but it's expensive for the user in the long term. We are using a 45mAh battery at the moment (might do a custom one in the future).
Power consumption details:
BLE base part:Sleep current = ~1uA
Fast advertisement mode (not connected to a master) = ~1.2mA
Normal operation (connected to a master and transmitting at regular intervals)= 300-600uA
The last one depends on the actual type of sensor/actuator, some of them are transmitting more frequently than others.
Top sensing acting part: this can range from hundreds of nA for buttons and low power sensors and can go up to tens of mA for motors - although motors have a larger battery ofcourse.
The software is at the moment point to point but we are developing the mesh version of it now. It will be ready by March when we ship.