Andy Rubin and his 60Ghz USB 3.0 connectors

[Garrett Kinsman]

Andy just announced a chip based 60Ghz USB 3.0 magnetic dock for his phone. https://www.youtube.com/watch?v=HCjjnVNXDLI

Viewing the previous silicon iterations of the ARA dock, it seems that the new generation of THz chips may be much more efficient. Using USB 3.0 interface, and the lowering cost of those chipsets are making ARA more and more feasible.

Not sure how this impacts battery life, or having interference from 12 60ghz transmitters. What I do know is that devices like iPhone use USB between internal "Modules" on the PCB.

I know this forum is pretty quiet these days, but ARA becomes more feasible each day. 

[Sajarin]

How possible is it for the hobbyists and developers here on this forum to come together and finish building out what Google Ara started? 

[Tim Southwick]

From a technical standpoint, I don't see any reason not to.

From a legal standpoint, I'm not a lawyer, but I'd guess that anything developed using Ara's specifications would need to be somehow licensed for use with Ara at some point, and any sold hardware would need to have some amount payed to them; effectively paying Google what they would have been paid for a license if it was officially launched. Again, I'm not a lawyer.

On the other hand, if you wanted to create a new specification, then as long as it doesn't use anything from Ara then you should be fine legally. Again, I'm not a lawyer, so I could be wrong.

[Garrett Kinsman]

Hey guys!

Glad to hear that some people still have ARA in mind. I think the best way forward would be to license the designs from google, and bankroll the final development and production with some sort of revenue share. Its not impossible by any means, just very expensive. I think they did a lot of valuable progress, and starting from scratch would be a huge waste. I hear rumors of sad ARA engineers in Lyft rides. The four main problems I understand were 1. Power consumption, 2. Software, competition with Pixel, and 3. money money money.

I think the most useful aspect for ARA would be to design a device with removable screen and add a large battery. This would become a new kind of computing that could be deployed for industrial IoT applications and beyond. I think a lot of people are afraid of hooking their critical systems to Windows these days (Referencing last weeks source code leak) , if there was a modular reference device for Android it would be very useful! 

I was in the Himalayas last week speaking with WISP people that would love a hyper configurable ethernet router with battery backup. What ARA could become if we remove the screen gets very interesting.

It is a shame that Google decided to copy the iPhone with their own device, and not radically innovate what it means to be a computing device. The Pixel is cool and takes probably the best photos, but it is still "Inspired" by the iPhone. Every smartphone today copies the iPhone model, we need something new.

PS I wrote a medium article on IoT if anyone is interested https://medium.com/@garrettkinsman/the-coming-wave-in-iot-and-how-it-will-change-the-world-forever-6f8eddab0501

[m d]

I've kept the Ara meetup group alive in SF for just this very reason. Will write up more, but in general, want to do a modular mobile device ecosystem with components using wireless connectivity. This would include modular CPU and mesh networking unit, touch screen display unit like a typical smart phone, augmented reality display, wrist display, cameras, ear, and microphone units, and many types sensor units.

The overall idea is to unbundle mobile into a distributed system.

> --
> You received this message because you are subscribed to the Google Groups "Ara Module Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to ara-module-devel...@googlegroups.com <mailto:ara-module-devel...@googlegroups.com>.
> To view this discussion on the web visit https://groups.google.com/d/msgid/ara-module-developers/8a637cc9-3993-4606-a5f5-201941c871ba%40googlegroups.com <https://groups.google.com/d/msgid/ara-module-developers/8a637cc9-3993-4606-a5f5-201941c871ba%40googlegroups.com?utm_medium=email&utm_source=footer>.
> For more options, visit https://groups.google.com/d/optout.
>
> <http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient> Virus-free. www.avg.com <http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient>
>
> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>


--
to connect on Linkedin,
http://www.linkedin.com/inviteFromProfile?from=profile&key=15253674&firstName=Mike&lastName=Doty

[Sajarin]

I think we should form a group on Slack or something. 

If we can organize ourselves in terms of expertise, I believe we can finish what Ara started. 

We'll get lawyers, hardware engineers, designers and software developers. A lot of people wanted Google Ara. 

What do you guys think? 

[m d]

yes, agree with setting up slack

> --
> You received this message because you are subscribed to the Google Groups "Ara Module Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to ara-module-devel...@googlegroups.com <mailto:ara-module-devel...@googlegroups.com>.

> To view this discussion on the web visit https://groups.google.com/d/msgid/ara-module-developers/b1ebc251-0899-4261-bf4f-37f07751e833%40googlegroups.com <https://groups.google.com/d/msgid/ara-module-developers/b1ebc251-0899-4261-bf4f-37f07751e833%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Garrett Kinsman]

This is a great idea. I was thinking we should all meet up in SF, except I'm living in india!

Who knows how to setup slack groups?

[m d]

We can set up a channel for starters in the Ark crypto slack. This is paid, so posts won't get deleted. Then if group gets large we can set up its own slack. We have over 3000 members in the public Ark slack, 60 in the private one.

On 6/26/2017 10:32 AM, Garrett Kinsman wrote:
> This is a great idea. I was thinking we should all meet up in SF, except I'm living in india!
>
> Who knows how to setup slack groups?
>

[Sajarin]

This is fantastic idea. As long as Ark Crypto doesn't mind/care. 

Everyone here should drop their emails so Mike can add them. 

my email is sajari...@gmail.com 

[Tim Southwick]

Sounds great :) Once it's set up, I'd recommend setting up a dedicated thread to a) announce that the group is set up in Slack, and b) summarize our progress every once in a while, to show that progress is still being made. I've almost got my masters degree, and plan to join once I've had a bit of experience in the work space. I've had a fair amount of experience as a hobbyist, but want to have more experience with lower levels of abstraction, and some experience with production if possible.

[Sajarin]

If no one is willing to create the slack group, I would be more than happy to start it off. I was hoping we could use the Ark Crypto Slack that Mike mentioned, but he hasn't replied back thus far. 

[m d]

yes, still want to do it.

On 6/28/2017 9:53 AM, Sajarin wrote:
> If no one is willing to create the slack group, I would be more than happy to start it off. I was hoping we could use the Ark Crypto Slack that Mike mentioned, but he hasn't replied back thus far.
>

> --
> You received this message because you are subscribed to the Google Groups "Ara Module Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to ara-module-devel...@googlegroups.com <mailto:ara-module-devel...@googlegroups.com>.

> To view this discussion on the web visit https://groups.google.com/d/msgid/ara-module-developers/7fb6dd1d-d50b-4a51-8f5f-61e4eaef4493%40googlegroups.com <https://groups.google.com/d/msgid/ara-module-developers/7fb6dd1d-d50b-4a51-8f5f-61e4eaef4493%40googlegroups.com?utm_medium=email&utm_source=footer>.

> For more options, visit https://groups.google.com/d/optout.
>
> <http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient> Virus-free. www.avg.com <http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient>
>
> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

[Sajarin]

Anyone still up for this? I can set up a slack group within 24 hours if need be. I really want to finish what Ara started. 

[Greg KH]

On Sun, Jul 16, 2017 at 02:26:00PM -0700, Sajarin wrote:
> Anyone still up for this? I can set up a slack group within 24 hours if need
> be. I really want to finish what Ara started. 

I kind of doubt that the lack of a slack group is what is holding anyone
back here. Mailing lists work just fine...

good luck!

greg k-h

[Garrett Kinsman]

Everyone,

I'm fine either way. Personally I kind of hate slack. Their $9B valuation is everything that is wrong with the valley. Maybe we keep it here until we actually start building, as this is the OG channel for everything ARA. -maybe new people will check in and join.

Let's start with the main issues with ARA. What do you guys think it was? Please add and contribute your thoughts, this will give us a good "Stat of the union" and help realize our next steps.

-Politics at google in relation to Pixel project (my main guess) Why innovate when you can just copy Apple?

-Battery life issues because of modular architecture

-Software issues in Android

-Politics between forking android to support ARA (another likely cantidate)

-BOM cost issues for final release 

Personally, I'm pretty much broke and jobless, so have to get that one figured out first. But I'm happy to work on this any way I can. I'm screwing around with IoT stuff, and ARA becomes more valuable each day. 

GK

[Tim Southwick]

Fair warning, I was feeling tired when I wrote this and didn't proofread what I wrote as well as I normally would.

Some problems I found with Ara:

• Space within each module - making the modules connect to a frame magnetically was dropped mainly due to space limitations. You could get more space by making the plastic thinner, but then the modules are less durable.
• New technology! There was plenty of technology being developed as part of Ara that would theoretically live on even if Ara didn't succeed. Here's what I remember out of it, although I am not sure how any of them are doing today:
  
• 3D printing - allow production of module covers that could be customized with images, transparency, probably some amount of depth.
• Tightly integrated hardware development environment - the idea was to make hardware nearly as easy to develop as software - you design it, everything you could want to test could be tested in a single development environment (temperature, radio emissions, electrical characteristics...). This would make hardware development much simpler to develop - but there's a TON of work involved in developing that sort of system. I haven't got a clue as to how far along the development of such a development environment got.
• Ambition - it's great that Project Ara was ambitious, and it's because of that ambition that it would have been so revolutionary. They got a working prototype, and showed that the camera could be taken out and put back in without rebooting. However, they wanted to get a lot done in a very short space of time, planning having a prototype available for sale more than once (which never got released). However, it might have been wiser to scale back their plans for the first iteration of Ara. Here's some examples:
• Instead of 3D printing custom module covers, use existing mold techniques, allowing a small range of colors to be selected from in each of the module sizes. Make the covers replaceable, so that the colors can be changed, and more customized covers could be swapped in at a later date.
• Instead of trying to have a set of high end parts, low end parts, and everything in between available at launch from Google, just make some middle of the road parts that are compatible, affordable for most users, and powerful enough to not be too frustrating. The parts are replaceable, so a faster processor could be released some months later for those who need it.
• While the Ara forum have been very good at coming up with ideas that would only work on a system like Ara, like how to react to multiple processor modules, those kinds of special features don't need to be thought out at the beginning completely. There needs to be enough thought going into it that the hardware would be compatible, or at least be able to recover and prevent damage, but the software can be changed. A new version of the Ara code could enable support for new combinations of modules, and new modules could contain both a link to where to get their latest code, and a fall-back set of code if that's not available.
• There seemed to be an insistence on having unique hardware available at launch; sure, having the hardware at launch would look great for Ara, but I don't think it would be a problem for Ara if it weren't available right away. Having a basic Ara phone available would be very useful for developing hardware, as the specification for the hardware isn't in development, but fixed in place for that one iteration of Ara. Develop a simple module with an LED and little else? If it works on the first Ara phone to be released to the public, then you can create more advanced devices and not need to worry about the interface changing. During development? Well, the connector changed from short range wireless to a physical connector late in Ara's life, and I wouldn't be surprised if the chip handling the network needed to change if development resumed.
  

Not a problem Ara had, but I'm considering the state of USB 3.1 Type-C. It handles enough power to charge a laptop, high resolution video at various frame rates, ethernet, and of course, USB, all at high speeds. It also supports some alternate modes that can be used for, say PCIe or analog audio. This sounds ideal for something like Ara, but it sounds like there isn't currently a chip that handles all of the features of USB 3.1 Type-C yet, and which ports support which features is currently rather poorly communicated. I think USB 3.1 Type-C is a good technology, and could be very useful for a future version of Ara, but isn't currently in a state I'd recommend for all of its features at once.

[Garrett Kinsman]

Good points!

Additive MFG: The ARA Team quickly abandoned 3D printing as costs would be way too high. Like 10x of injection molding. I also met with some of the 3d systems guys at the Dev Con I, and they were huge jerks. (Albeit I was in high school). 

Space: Space was a big issue, but one I think iterations could fix. Bulky phones don't sell well, and you need the economies of scale to bring cost down. I think positioning ARA as an industrial computer / IoT platform  may have been more interesting. I think back to the Iridium satellite phone debacle, Motorola couldn't seem to find the "Business traveler" market fit, and they were preparing to de-orbit billions of dollars of satellites. ARA suffers the same issue, once Motorola realized that the market fit was really government, they has sold the system for pennies on the dollar, and perhaps the most complex system ever built was saved- and is still in operation. 

Custom SOC: I do remember in regards to the space issue, ARA was planning on a custom System on a Chip. These are ridiculously expensive, and would have been required to save space and power for each module. This may have been an economic barrier to mass production.

USB C: Many people don't know this, but USB C leaks 2.4 GHz!! Any RF transmitters should be about 10-12cm away from any USB-C interface. There are workarounds but this adds complexity! I'd love to see teardowns and RF tests of Andy Rubin's arrangement. I do know they have ex-ATAP people on their team. 

Software: I don't know much on this issue, except that Android may have had to have been heavily modified. 

Economics: Ara team showed off a V2 mockup that was considerably thinner. Perhaps the break-even projections to make that version sustainable were too high or unreasonable. Another issue is San Francisco is living in a shiny bubble of iPhone. APPLE IS GOD! SINNERS SHALL PERISH THEE FATES OF HTC. Where I'm living in Bangalore, the whole country runs Android and Windows XP. Siri doesn't really work, nor does Apple maps. it's like a huge joke.  I think there is a lot of demand for something like ARA, especially with a Chrome OS like desktop interface. 

Any other technical issues we can think of? Might be useful to contact the people at NK labs, the original team behind ARA.

I think it really just comes down to $$$$$$$$. We could probably negotiate a license of the IP from google, considering it would enhance their brand- but:

1. forking Android might be a huge problem

2. Custom SOCs for module interfaces might be cost prohibitive. Would love to see the advantages of ARA's custom interface vs something like USB. For example iPhone uses USB for their internal connections, and this might drastically lower costs.

I don't see any other huge technical roadblocks

GK

[m d]

have a lot to say but going to blockchain conference today, will reply later.

On 7/17/2017 11:10 PM, Garrett Kinsman wrote:
> Good points!
>
> *Additive MFG:* The ARA Team quickly abandoned 3D printing as costs would be way too high. Like 10x of injection molding. I also met with some of the 3d systems guys at the Dev Con I, and they were huge jerks. (Albeit I was in high school).
>
> *Space*: Space was a big issue, but one I think iterations could fix. Bulky phones don't sell well, and you need the economies of scale to bring cost down. I think positioning ARA as an industrial computer / IoT platform may have been more interesting. I think back to the Iridium satellite phone debacle, Motorola couldn't seem to find the "Business traveler" market fit, and they were preparing to de-orbit billions of dollars of satellites. ARA suffers the same issue, once Motorola realized that the market fit was really government, they has sold the system for pennies on the dollar, and perhaps the most complex system ever built was saved- and is still in operation.
>
> *Custom SOC*: I do remember in regards to the space issue, ARA was planning on a custom System on a Chip. These are ridiculously expensive, and would have been required to save space and power for each module. This may have been an economic barrier to mass production.
>
> *USB C*: Many people don't know this, but USB C leaks 2.4 GHz!! Any RF transmitters should be about 10-12cm away from any USB-C interface. There are workarounds but this adds complexity! I'd love to see teardowns and RF tests of Andy Rubin's arrangement. I do know they have ex-ATAP people on their team.
>
> *Software*: I don't know much on this issue, except that Android may have had to have been heavily modified.
>
> *Economics: *Ara team showed off a V2 mockup that was considerably thinner. Perhaps the break-even projections to make that version sustainable were too high or unreasonable. Another issue is San Francisco is living in a shiny bubble of iPhone. APPLE IS GOD! SINNERS SHALL PERISH THEE FATES OF HTC. Where I'm living in Bangalore, the whole country runs Android and Windows XP. Siri doesn't really work, nor does Apple maps. it's like a huge joke. I think there is a lot of demand for something like ARA, especially with a Chrome OS like desktop interface.

>
> Any other technical issues we can think of? Might be useful to contact the people at NK labs, the original team behind ARA.
>
> I think it really just comes down to $$$$$$$$. We could probably negotiate a license of the IP from google, considering it would enhance their brand- but:
> 1. forking Android might be a huge problem
> 2. Custom SOCs for module interfaces might be cost prohibitive. Would love to see the advantages of ARA's custom interface vs something like USB. For example iPhone uses USB for their internal connections, and this might drastically lower costs.
>
> I don't see any other huge technical roadblocks
>
> GK
>
>
> On Tuesday, July 18, 2017 at 6:32:41 AM UTC+5:30, Tim Southwick wrote:
>
> Fair warning, I was feeling tired when I wrote this and didn't proofread what I wrote as well as I normally would.
>
> Some problems I found with Ara:
>

> * Space within each module - making the modules connect to a frame magnetically was dropped mainly due to space limitations. You could get more space by making the plastic thinner, but then the modules are less durable.
> * New technology! There was plenty of technology being developed as part of Ara that would theoretically live on even if Ara didn't succeed. Here's what I remember out of it, although I am not sure how any of them are doing today:
> o 3D printing - allow production of module covers that could be customized with images, transparency, probably some amount of depth.
> o Tightly integrated hardware development environment - the idea was to make hardware nearly as easy to develop as software - you design it, everything you could want to test could be tested in a single development environment (temperature, radio emissions, electrical characteristics...). This would make hardware development much simpler to develop - but there's a TON of work involved in developing that sort of system. I haven't got a clue as to how far along the development of such a development environment got.
> * Ambition - it's great that Project Ara was ambitious, and it's because of that ambition that it would have been so revolutionary. They got a working prototype, and showed that the camera could be taken out and put back in without rebooting. However, they wanted to get a lot done in a very short space of time, planning having a prototype available for sale more than once (which never got released). However, it might have been wiser to scale back their plans for the first iteration of Ara. Here's some examples:
> o Instead of 3D printing custom module covers, use existing mold techniques, allowing a small range of colors to be selected from in each of the module sizes. Make the covers replaceable, so that the colors can be changed, and more customized covers could be swapped in at a later date.
> o Instead of trying to have a set of high end parts, low end parts, and everything in between available at launch from Google, just make some middle of the road parts that are compatible, affordable for most users, and powerful enough to not be too frustrating. The parts are replaceable, so a faster processor could be released some months later for those who need it.
> o While the Ara forum have been very good at coming up with ideas that would only work on a system like Ara, like how to react to multiple processor modules, those kinds of special features don't need to be thought out at the beginning completely. There needs to be enough thought going into it that the hardware would be compatible, or at least be able to recover and prevent damage, but the software can be changed. A new version of the Ara code could enable support for new combinations of modules, and new modules could contain both a link to where to get their latest code, and a fall-back set of code if that's not available.
> o There seemed to be an insistence on having unique hardware available at launch; sure, having the hardware at launch would look great for Ara, but I don't think it would be a problem for Ara if it weren't available right away. Having a basic Ara phone available would be very useful for developing hardware, as the specification for the hardware isn't in development, but fixed in place for that one iteration of Ara. Develop a simple module with an LED and little else? If it works on the first Ara phone to be released to the public, then you can create more advanced devices and not need to worry about the interface changing. During development? Well, the connector changed from short range wireless to a physical connector late in Ara's life, and I wouldn't be surprised if the chip handling the network needed to change if development resumed.

>
>
> Not a problem Ara had, but I'm considering the state of USB 3.1 Type-C. It handles enough power to charge a laptop, high resolution video at various frame rates, ethernet, and of course, USB, all at high speeds. It also supports some alternate modes that can be used for, say PCIe or analog audio. This sounds ideal for something like Ara, but it sounds like there isn't currently a chip that handles all of the features of USB 3.1 Type-C yet, and which ports support which features is currently rather poorly communicated. I think USB 3.1 Type-C is a good technology, and could be very useful for a future version of Ara, but isn't currently in a state I'd recommend for all of its features at once.
>
>
> On Monday, July 17, 2017 at 1:12:09 AM UTC-4, Garrett Kinsman wrote:
>
> Let's start with the main issues with ARA. What do you guys think it was? Please add and contribute your thoughts, this will give us a good "Stat of the union" and help realize our next steps.
>

> --
> You received this message because you are subscribed to the Google Groups "Ara Module Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to ara-module-devel...@googlegroups.com <mailto:ara-module-devel...@googlegroups.com>.

> To view this discussion on the web visit https://groups.google.com/d/msgid/ara-module-developers/6fbcfdf9-9757-49bd-a498-2f230e1a9efc%40googlegroups.com <https://groups.google.com/d/msgid/ara-module-developers/6fbcfdf9-9757-49bd-a498-2f230e1a9efc%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Garrett Kinsman]

Fyi:

Facebook’s secretive hardware division is exploring modular smartphones - The Verge
https://apple.news/A1SyAeQ7-S5eDq_h4oFLw6w

Dugan has roots in the Ara team

[Sajarin]

Nice find, i didn't expect this from Facebook. However, I suspect that it'll take years before anything concrete comes out. 

I'm liking the discussion thus far, and I think both Garrett and Tim have hit the nail on the head with some of the problems with Ara. 

Our next step should be getting more people involved in this project. I'm going to try posting this on the Project Ara subreddit and see what responses we get.

Other than that we should definitely be brainstorming on how to start. Getting a license from Google could take a long time. I'm not even sure how that process works.   

 

[Tim Southwick]

I just thought of something that would make a modular smartphone system more appealing to manufacturers, but it's going to sound a bit weird. If there were a system in place to get rights to the circuits inside of modules, it would be possible to make non-modular phones that are smaller and lighter, but because the circuits are identical to that of the modular parts it's made of, the non-modular phones could be updated far more easily. I'm thinking somewhat like open platforms like Arduino or Raspberry Pi in terms of physical scale and support. This would not require the hardware developers to provide full design details, either - ARM doesn't manufacture their production chips, but designs parts that can be put together and manufactured on a single chip by other companies.

[m d]

I've added a channel called "modular_mobile" to the Ark Cryptocurrency Project Slack for those interested in developing open source modular mobile devices. The URL to join the slack is https://ark.io/join-ark-slack/ . For those not familiar with Slack, to finf and join the channel after joining the Slack, please click on "channels" to see the list of channels, click "preview", and then click "join".

More than just cell phones, the future of mobile is likely to be personal networks of wireless devices. Open source modular mobile with personal wireless attached devices is part of my roadmap. Instead of having everything based in the phone, to unbundle it so the radios, CPU, storage, cameras, displays, mics, speakers, earbuds, sensors, and other devices can be distributed, like wrist displays, augmented reality glasses, PTZ cameras, antennas, and modular user configurable core module with storage, CPU, radios can be distributed as wearables and even have remote devices function as a personal mobile ecosystem.