Designing Hardware for the Internet of Things | Hacker News Reader

Designing Hardware for the Internet of Things

evenchick.com

38 pointsevck26 comments

Awesome - keep up the great work!

I'm working on one of the next upstream problems in the IoT space: once you have a way to connect a device this to a network, how do you tie it into the greater set of things, program it, make it discoverable, and manage it.

My stab at this is called Magenta, which is an MIT licensed project based on node.js that aims to make it easy to build devices that can securely communicate in real time and applications that interact with those devices over a message passing model (its hard to explain in one paragraph but think of it like this: twitter for devices with agents watching those messages).

I'm only a month into it and a few hundred commits, but if you are interested, check out the (very early) project at https://github.com/magentajs or contact me on twitter at @timpark. I would love to get some feedback or find folks interested in helping collaborate on one of the subprojects (client sdk, service, administration tool, or use it to build devices and provide feedback and bugs).

Ok, I'm done hijacking your thread!

Thanks for sharing that, I'll check it out. One problem I'm looking forward to working on is what the embedded device will do to take care of communications, and what the firmware to do that looks like. Lots of neat problem including authentication, minimizing power consumption, etc...

Similarly, check out plumgarage.com; tackling a different side of the problem.

Another interesting internet of things module: " The RS9113 M2MCombo supports dual-band 802.11n and dual-mode BT 4.0 in addition to Zigbee. Redpine (San Jose, Calif.) will ship the device before June in a variety of module form factors at a cost of $6 in 100,000-unit quantities."[1]

Hopefully it will also get into a shield form factor, at a reasonable price, since it seems ideal to trying all kinds of things out.

[1]http://www.eetimes.com/electronics-news/4406270/Redpine-devi...

The CC3000 is IPv4 only - what the fk?! I understand it greatly increases cost in manufacturing, but isn't one of the key sale points of IPv6 that the IPv4 address space is going to run out in part because of the Internet of Things!?

Eh. Is everything in my house (i.e. my subnet 192.168..) going to run out of IPV4 addresses? Right now cost is a far more important factor.

People need to get used to IPv6 development. IPv6 development and deployment is way way behind schedule already, to sell NEW stuff today without IPv6 support is foolish to say the least. All that stuff will have to be replaced in not-so-far future.

Selling stuff that can't handle IPv6 isn't foolish, it just makes you a leech on society, since everyone else will have to clean up your mess. Buying such equipment, on the other hand, is quite foolish.

FWIW, that image is a preview from the OSH Park ordering site[1]. I can highly recommend the service for prototype 2- and 4-layer boards. The price is right, the turnaround time and customer service are awesome, and the specs are better than anything else I've found in the price range. And they're purple.

[1] http://www.oshpark.com

Yep, it's an OSHPark render. I'm not quite ready to release the design files, so I figured I'd post that picture since it's nice an purple.

I'd also recommend OSHPark.

Good luck with the waveguide! Have you done anything that high-speed with OSH Park before? I'm waiting on a PCIe board from them that I hope will do Gen3 speeds (8GHz) over 2.5-inch traces, but the lack of impedance control has me nervous.

I suppose the waveguide looks pretty short (under an inch?) so it might not matter much?

Haven't done anything high speed, but I'm not too worried since it's short. It might not be perfect on the first run, but hopefully it'll work.

8 GHz and multiple traces would be a bit more concerning without impedance control. Hope it works out.

To be clear, my differential pairs are impedance controlled in the sense that their trace/space yield the correct single-ended and differential impedance values at the nominal dielectric permittivity and thickness, but OSH Park doesn't additionally guarantee permittivity and thickness tolerances to guarantee that my nominally-50-ohm traces will be actually-50-ohm traces +/- 10%. That said, I've run the numbers (using the excellent and free Saturn tools [1]) for the stated Er range of 4.2 +/- 0.3, and a little variation in the nominal 7.8 mil thickness, and things still looked OK.

Fingers crossed!

[1] http://www.saturnpcb.com/pcb_toolkit.htm

I'll point out that it is very clear that there will be BLE support in stock android(Samsung supports BLE via a device specific SDK) after the Google IO release of the Android.

Google's Glass supports BLE based on the FCC approval documents. It appears that BLE support was prioritized after the Bluetooth stack rewrite that happened in 4.2.

The Galaxy Nexus, Nexus 4, Nexus 7 and Nexus 10 all have hardware capable of BLE that should be available after the update.

Great work. I've been keeping a close eye on the CC3000 for some time. I've also looked at the Wi2Wi modules/chips. I've been working (whenever work and home allows) on some hardware to connect the CC3000 with a Xilinx FPGA. Hopefull, I will have some results soon.

Also, I really like seeing hardware on the front page.

I've played around with a WiFly module sold by sparkfun that seems pretty functional and reasonably cheap at 35$.

https://www.sparkfun.com/products/10822

Any idea how much cheaper this could be?

By the way your site doesn't seem to have a valid RSS or Atom feed. Your html has an incorrect feed link back to the homepage itself:

  <link href="./" type="application/atom+xml" rel="alternate" title="Eric Evenchick ATOM Feed" />

This looks like a nice shield for that cc3300. Nice to see more hw-related stuff on HN - thanks for sharing.

I'd like to write more hw stuff for the HN audience. I'm curious if there's anything in particular that the HN community, who seem to be more sw focused, would like to see.

IMHO most real innovation is happening in the space between hw and sw (duino / pi / etc.). The idea that you can call yourself an "innovator" and only know software is quickly becoming quaint. I don't claim to speak for the HN community-at-large, but I'm interested in seeing anything in the vein of hw/sw synthesis like this.

Real innovation is happening everywhere. Unfortunately, only the consumer-facing (and hobbyist) stuff gets covered here. Arduino, Raspi, etc. are interesting, but not technologically innovative. Their innovation is in addressing poorly-served markets: just as important, but not quite the same thing.

The "internet of things" is interesting from a consumer perspective. I've been involved in it for a few years in terms of being able to remotely monitor and diagnose complex equipment. There's a long list of companies with lots of dollars behind them that want to get started in this arena and don't know how. I suspect that this is soon going to be a hot area of consulting.

Just look at what businesses like Axeda have been doing for the last 10 years if you want to see where industry is heading.

> IMHO most real innovation is happening in the space between hw and sw (duino / pi / etc.). The idea that you can call yourself an "innovator" and only know software is quickly becoming quaint.

On what basis do you make this claim? I don't think we've come close to exhausting the space of useful and interesting things that can be done in pure software on general-purpose computers.

First, "IMHO" means "In my honest opinion"..so I wasn't claiming it as fact. But since you asked, the writing I'm seeing on the wall is a misto of the declining quality of sw startup ideas, the burgeoning of hw startups and the common sense to recognize that we are beating the dead horse of this one mode of sense and actuation a little too long..even PG said that he is looking more into hw (sorry no link).

I'm a software guy; I wish it wasn't true, because it would be a lot more convenient for me - but I'm not going to pretend it isn't happening...I'm learning as much pi/duino/tronics as quickly as I can.

Nice. Maybe my project to set up a bunch of open source toolchains for firmware development would be interesting then. Thanks for the feedback!

Fear not, there are hardware lovers here too.

Cool project.

If your interested in some feedback on the PCB layout I'd be happy to help (e-mail in profile). There are a few things about the pictured layout that will make it very prone to noise and EM radiation as well as some heat dissipation issues.

I would really appreciate your feedback. Can't see your email though. Could you get in touch through my site (under About)?

Reply on news.ycombinator.com