Some of you who follow me on twitter may have noticed that from time to time I have mentioned various things relating to an IoT (for those not up with the lingo Internet of Things) device. I figure it is time to finally break some silence on this.
Before I jump into the rest of the post, there is no major reveal and no demonstrations included. I am still very much in the development phase on the project and do not have anything that can be shown. As a result the full project is still going to remain an Internet of Secret Things device. I shall however document the process of building the prototypes to share my experience with all of you :)
The device itself isn't a single device but rather a collaboration of several devices working together for the same end task, ending with a network of things connected to the internet of things. The easiest way to look at it is a control device and an interface device.
These little devices are the devices that do the actual work. They aren't much more than an MCU, a network bus and circuitry to do a given function. These device can be broken into 2 parts, one part to control the actual function, and a generic main board. The main board houses the MCU and network communications (by ZigBee, sub-GHz RF, wireline) hardware. The second board contains controllers and other circuitry to do the defined function of the accessory. These systems may be offered to other OEMs to intgrate connectivity into their devices without compromising the security of the network of devices.
These devices are planned to run FreeRTOS and remain an open platform providing an open API for third parties to be able to extend services provided, although some security functions would be unavailable to protect other devices.
The current design is built around the Atmel range of cortex MCU's future redesigns will be made to reduce the currently high (US$30-40) price of building a board. Prototypes have not yet been produced but parts have started arriving.
2 Atmel SA3X4C for prototype.
2 Atmel AES EEPROMs for security functions.
Special thanks to Atmel for the samples.
The main part of the device is the interface device. This acts as an interface between the control devices and you. This unit is one complete unit and features a touchscreen as well as full network connectivity. While control devices live on a separate network, the interface device acts as a gateway between the networks. Interface devices run on an Arm Cortex-A5 SoC for performance at a lower power profile. These devices will have the ability to communicate with each other if paired to mesh together many accessories under universal control.
The interface devices are planned to be more closed in nature due to the requirement of a more advanced OS to run the system. An RTOS such as QNX or vxworks are required to keep the boot time to a minimum and responsiveness high.
2 Atmel SAMA5 SoC's
One very tiny thermal probe.
I would like to thank Atmel for the samples that arrived today after a very short 3 business day shipping from central USA to Australia. Takes the record for the quickest shipment from the USA to AU I have ever seen.
So what next?
Next step is to get the prototype boards I have been designing and bring up the RTOS' chosen up on these boards. Following that a test implementation and demonstration. Sounds easy but it is a lot more work than the very brief description suggests.
I have been left several times with the question in my head "To kickstarter or not to kickstarter?". This has been a very difficult question to answer, because I would very much love to open up to kickstarter, reveal all device functions and see how the general community responds, but alas I do not have any pretty artistic renders, no working prototypes and find the environment around kickstarter these days is very much based on the rewards offered for contributing certain amounts.
Things like offering T-Shirts for this much, merchandise packs for that much, finished products for this much etc. does not interest me, the first two seem like money has to be spent on making these items which could be better spent on improving the product being delivered.
Second issue that I have with kickstarter is being able to demonstrate working prototypes, videos and having pretty artists impressions to make your page look appealing. I do not have the resources to produce artists impressions, videos are a drain while still developing the project, and for applications like this most of the cost is in producing the prototype unit, once the prototype has been produced in a working state, production is just a stones throw away, and preorder type systems would be just as effective.
Just as an example the following is what I have left to go before a prototype can be finished:
- Prototype PCB production.
- New hot air solder station to replace aging equipment that is no longer capable of SMT.
- SMT resistors/capacitors/etc.
- Licenses for RTOSes.
These 4 items add up easily to over $1000. However I have already made relationships with supply and manufacturing in the process of getting to where I am now. Thus causing my issues with kickstarter based funding.
Shameless self promotion
Even with my hate of the reward base of services like kickstarter I am going to offer my one and only reward based offer. All donations received from today will be contributed towards this project. Your email from the donation (or if you put an alternative email in the notes) will be entered into a database of which when my devices are completed and offered for sale you will be notified, and offered a 10% lifetime discount. No minimum donation required, no minimum goal needing to be met.