[HH] the missing security camera revolution

[Drew Van Zandt]

 What yearly volume are you looking at?  Under 10k units, an $80 production
cost device should sell for $250 - $500 if you want to actually be able to
have a company that can support it.  BOM and production costs are generally
not the major contributors to the sale price of low-volume devices.

Using cheap Androids would make sense if you could replace the OS, except
of course that replacing the OS costs a great deal.

>From a post by Gui Cavalcanti, Asylum founder and a teacher @ Olin:


A random, interesting data point for you all -
[<https://www.facebook.com/myles.cooper.39>
redacted], a student in my Fundamentals of Business and Entrepreneurship
class at Olin, and his team have been working on a business plan that
essentially amounts to one person owning a MakerBotReplicator 2 and
printing products for public sale. They went in to the exercise wanting to
test the following hypothesis: since the MakerBot material is so cheap, and
the printer itself is so cheap, you should be able to make a business that
provides cheap 3D printed parts (i.e., much less expensive than Shapeways
or RedEye) to the public.

They found that if they assumed a barely-living wage for one person living
in Somerville, and if that person worked 10 hour days with the printer
running 24/7 (using an example 3D printed product they designed for the
exercise), they'd have to charge $15 per cubic inch of material just to
break even. If the printer wasn't running constantly, or if it ever had any
significant downtime, they'd have to charge more. By comparison, the actual
cost of the material is only $1.11 per cubic inch if you buy a 1kg spool
from MakerBot themselves.

Thought that might help put typical 3D printing costs in perspective for
people. It's only a cheap process if your time is free!

Similar rules apply to selling/making hardware, only more so.  If you are
not factoring in the cost of R&D, salaries, support, certification,
inventory maintenance, rent, insurance, marketing, etc. then well.... I
recommend you stay far away from actually trying this.  Distributors won't
carry your stuff unless they get a decent cut (20%+, usually, unless your
volumes are high), either.



*Drew Van Zandt*
Artisan's Asylum Board of Directors



On Mon, Apr 14, 2014 at 7:02 PM, Tom Metro <tmetro+hhacking at gmail.com>wrote:

> I receive ads from 123securityproducts.com. They seem to be typical of
> sites selling security cameras, although I only rarely look at the other
> sites, as I haven't been actively shopping for equipment in this space
> for a few years.
>
> The vast majority of the products they advertise are still analog
> cameras, usually sold in multi-camera bundles, with a central DVR
> (running some buggy Linux firmware, often with no source code provided).
>
> The ad I received today was one of the rare few promoting an actual IP
> camera. It has better resolution than you typically see. Instead of
> merely VGA or QVGA, its a 2.43 MP sensor, with a fixed lens. (A 2000 x
> 1241 pixel sensor, outputting 1920 x 1080 using H.264 at 30 FPS.) It
> also supports Power-over-Ethernet, which is good, but they want over
> $300 for this:
>
> https://www.123securityproducts.com/knc-hdi47b37.html
>
> I get that pro-grade hardware carries a premium, but the capabilities of
> this camera are likely blown away by a sub-$50 Android smartphone, which
> likely comes with a 3 MP camera, plus has WiFi and GSM radios.
>
> It seems easy to imagine how one could take a low-end smartphone SoC,
> combined with a high volume 5 MP camera module made for phones, and add
> a wired Ethernet port with PoE and some IR LEDS for night illumination,
> with production costs under $100, and a retail price of around $100.
> Replace Android with a very minimalist Linux or other open source RTOS
> optimized for reliability, and you'd have a great camera platform.
>
> Why haven't we seen this? Or is it out there, but I just haven't ran
> across its?
>
> Has anyone ran across any communities dealing with creating custom third
> party firmware for consumer or pro-sumer grade IP cameras? (Thee are
> plenty that run Linux and many that even provide source, but I haven't
> seem any communities form around hacking camera firmware.)
>
> Directly using these $50 smartphones is another option. There are apps
> to turn a phone into a security camera. But aside from the lens angle
> not necessarily being optimal for a security monitoring application, to
> make it a serious contender as a security camera you'd really need to
> replace Android with something more reliable, or at least make sure the
> phone has a good hardware watchdog, and customize the watchdog code to
> validate that all the functions critical to the camera are operational.
> (It would still be less than ideal due to using WiFi for networking and
> requiring local power and not having a mounting bracket.)
>
>  -Tom
> _______________________________________________
> Hardwarehacking mailing list
> Hardwarehacking at blu.org
> http://lists.blu.org/mailman/listinfo/hardwarehacking
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.blu.org/pipermail/hardwarehacking/attachments/20140414/a9f7c557/attachment.html>