Lots of interesting Internet-of-Things projects popping up lately.
17 views
Skip to first unread message
Luke Weston
May 18, 2012, 2:08:49 AM5/18/12
to connected-commu...@googlegroups.com
Interestingly, all these devices seem to be based around connectivity to 802.11 WiFi, which is maybe a good choice compared to other options such as ZigBee.
WiFi has traditionally meant relatively high hardware costs, but it gets cheaper and cheaper all the time. (For internet-of-things devices to be ubiquitous, the hardware devices have to be scalable deployable at a relatively low cost.)
The Microchip MRF24WB0MA, which is a pretty common choice for many systems where an AVR or PIC or ARM or similar microcontroller is connected to 802.11 WiFi in basic embedded projects, including some Arduino WiFi shields I've seen, is about $26 from DigiKey, which is cheaper than I've found them before.
So it is starting to get cheaper and relatively attractive, compared to other possible candidates such as MaxStream XBee modules.
WiFi has traditionally meant relatively high hardware costs, but it gets cheaper and cheaper all the time. (For internet-of-things devices to be ubiquitous, the hardware devices have to be scalable deployable at a relatively low cost.)
The Microchip MRF24WB0MA, which is a pretty common choice for many systems where an AVR or PIC or ARM or similar microcontroller is connected to 802.11 WiFi in basic embedded projects, including some Arduino WiFi shields I've seen, is about $26 from DigiKey, which is cheaper than I've found them before.
So it is starting to get cheaper and relatively attractive, compared to other possible candidates such as MaxStream XBee modules.
WiFi is tempting, because the device talks directly - but wirelessly - to the LAN, and you don't need to have a separate piece of hardware bridging across from a ZigBee network or some other sort of radio network like that across to ethernet or WiFi. Using something like a DS2401 1-Wire serial number chip to seed a unique MAC in every device makes this even easier to deploy scalably.
And, of course, essentially every home or business already has 802.11 WiFi infrastructure in place, so you're using that existing infrastructure and you don't have to go and have the cost of building that in order to communicate with your devices.
And, of course, essentially every home or business already has 802.11 WiFi infrastructure in place, so you're using that existing infrastructure and you don't have to go and have the cost of building that in order to communicate with your devices.
The "ReaDIYMate" basically appears to consist of an Arduino-compatible AVR, an SD card, an MP3 playback/decoder chip, a connector for a speaker and a servo, and an 802.11 radio. A nice platform for creative arty projects.
http://www.electricimp.com/
The Electric Imp looks very nice, very nicely engineered, but I fear it may be let down slightly by either lack of openness or excessively high price.
(The whole brains of the operation, the 32-bit microcontroller and the 802.11 radio, is inside the bit that has the same package as an SD card. Power supplies and I/Os are connected via an external board with an SD socket that the Electric Imp plugs in to.)
George Patterson
May 18, 2012, 2:39:49 AM5/18/12
to connected-commu...@googlegroups.com
On 18 May 2012 16:08, Luke Weston <reindeer...@gmail.com> wrote:
>
> http://www.kickstarter.com/projects/daisyworks/internet-your-thing-0
>
> Interestingly, all these devices seem to be based around connectivity to
> 802.11 WiFi, which is maybe a good choice compared to other options such as
> ZigBee.
>
> WiFi has traditionally meant relatively high hardware costs, but it gets
> cheaper and cheaper all the time. (For internet-of-things devices to be
> ubiquitous, the hardware devices have to be scalable deployable at a
> relatively low cost.)
>
> The Microchip MRF24WB0MA, which is a pretty common choice for many systems
> where an AVR or PIC or ARM or similar microcontroller is connected to 802.11
> WiFi in basic embedded projects, including some Arduino WiFi shields I've
> seen, is about $26 from DigiKey, which is cheaper than I've found them
> before.
>
> So it is starting to get cheaper and relatively attractive, compared to
> other possible candidates such as MaxStream XBee modules.
>
> WiFi is tempting, because the device talks directly - but wirelessly - to
> the LAN, and you don't need to have a separate piece of hardware bridging
> across from a ZigBee network or some other sort of radio network like that
> across to ethernet or WiFi. Using something like a DS2401 1-Wire serial
> number chip to seed a unique MAC in every device makes this even easier to
> deploy scalably.
>
So explain to me how it connects to the yoghurt again?
>
> http://www.kickstarter.com/projects/daisyworks/internet-your-thing-0
>
> Interestingly, all these devices seem to be based around connectivity to
> 802.11 WiFi, which is maybe a good choice compared to other options such as
> ZigBee.
>
> WiFi has traditionally meant relatively high hardware costs, but it gets
> cheaper and cheaper all the time. (For internet-of-things devices to be
> ubiquitous, the hardware devices have to be scalable deployable at a
> relatively low cost.)
>
> The Microchip MRF24WB0MA, which is a pretty common choice for many systems
> where an AVR or PIC or ARM or similar microcontroller is connected to 802.11
> WiFi in basic embedded projects, including some Arduino WiFi shields I've
> seen, is about $26 from DigiKey, which is cheaper than I've found them
> before.
>
> So it is starting to get cheaper and relatively attractive, compared to
> other possible candidates such as MaxStream XBee modules.
>
> WiFi is tempting, because the device talks directly - but wirelessly - to
> the LAN, and you don't need to have a separate piece of hardware bridging
> across from a ZigBee network or some other sort of radio network like that
> across to ethernet or WiFi. Using something like a DS2401 1-Wire serial
> number chip to seed a unique MAC in every device makes this even easier to
> deploy scalably.
>
Though seriously, If you are using I2C already in your project or you
have the pins free, it might be better to use a Microchip 24AA02E48,
it appears to be cheaper by a margin that could make it worth your
while to tweak the project design a little. It's not available in DIP
but uses just eight legs, not all of those are required (see datasheet
for exceptions as it depends on the package type) so dead bug or DIY
breakout board should be easy enough.
For the Laser Tag, project I have been considering using Unique ID
chips for the clips/magazines as it's rather cool to have the option
of passing physical tokens representing ammunition. But perhaps not in
the first pass, need comms first. :-P
Regards
George
Andy Gelme
May 18, 2012, 3:59:14 AM5/18/12
to connected-commu...@googlegroups.com
hi All,
On 2012-05-18 16:08 , Luke Weston wrote:
> Interestingly, all these devices seem to be based around connectivity
> to 802.11 WiFi, which is maybe a good choice compared to other options
> such as ZigBee.
Depends upon your application. There are other considerations, such as
network topology, power consumption, sleep/wake-up/connect cycle ... and
the amount of software stack required by TCP/IP, DHCP, etc.
When it comes to pushing the boundaries with WiFi ... at least 7 years
ago, the Eye-Fi (WiFi SD card) was already showing what was possible
with small form-factor WiFi ... in devices other than laptops (a radical
idea at the time).
> WiFi is tempting, because the device talks directly - but wirelessly -
> to the LAN, and you don't need to have a separate piece of hardware
> bridging across from a ZigBee network or some other sort of radio
> network like that across to ethernet or WiFi.
A WiFi router is a separate piece of hardware bridging from the WiFi
network to the Ethernet.
> And, of course, essentially every home or business already has 802.11
> WiFi infrastructure in place, so you're using that existing
> infrastructure and you don't have to go and have the cost of building
> that in order to communicate with your devices.
That certainly gives WiFi an edge. People will use it because of
familiarity and ubiquity (which by themselves aren't the best criteria
for choosing a technology).
However, it would be very prudent to run a separate network for your
building automation ... and not the (wireless) LAN that you are using
and sharing with everyone else who drops by.
- - - - - - - - - - - - - - - - - -
The big problem for any permanent network (deployed and in use for
years), such as building automation, is that the majority of devices
will (or should) have cabling for power. Constantly changing batteries
in a large variety of devices over time will be problematic. So, if you
have cabling for power, you might as well use that cabling for
networking as well, e.g PoE. Cable is typically faster, cheaper, more
reliable and more secure than wireless ... with the major disadvantage
being that it is difficult/costly to retro-fit into existing buildings.
In contrast, the problem for 2.4 Ghz WiFi is the crowded and unlicensed
spectrum.
As for wireless networking ... I anticipate that there will continue to
be a number of complimentary/competing technologies, e.g WiFi,
Bluetooth, ZigBee, ZWave and even generic 433 Mhz radio for some time to
come.
--
-O- cheers = /\ /\/ /) `/ =
--O -- http://www.geekscape.org --
OOO -- an...@geekscape.org -- http://twitter.com/geekscape --
On 2012-05-18 16:08 , Luke Weston wrote:
> Interestingly, all these devices seem to be based around connectivity
> to 802.11 WiFi, which is maybe a good choice compared to other options
> such as ZigBee.
network topology, power consumption, sleep/wake-up/connect cycle ... and
the amount of software stack required by TCP/IP, DHCP, etc.
When it comes to pushing the boundaries with WiFi ... at least 7 years
ago, the Eye-Fi (WiFi SD card) was already showing what was possible
with small form-factor WiFi ... in devices other than laptops (a radical
idea at the time).
> WiFi is tempting, because the device talks directly - but wirelessly -
> to the LAN, and you don't need to have a separate piece of hardware
> bridging across from a ZigBee network or some other sort of radio
> network like that across to ethernet or WiFi.
network to the Ethernet.
> And, of course, essentially every home or business already has 802.11
> WiFi infrastructure in place, so you're using that existing
> infrastructure and you don't have to go and have the cost of building
> that in order to communicate with your devices.
familiarity and ubiquity (which by themselves aren't the best criteria
for choosing a technology).
However, it would be very prudent to run a separate network for your
building automation ... and not the (wireless) LAN that you are using
and sharing with everyone else who drops by.
- - - - - - - - - - - - - - - - - -
The big problem for any permanent network (deployed and in use for
years), such as building automation, is that the majority of devices
will (or should) have cabling for power. Constantly changing batteries
in a large variety of devices over time will be problematic. So, if you
have cabling for power, you might as well use that cabling for
networking as well, e.g PoE. Cable is typically faster, cheaper, more
reliable and more secure than wireless ... with the major disadvantage
being that it is difficult/costly to retro-fit into existing buildings.
In contrast, the problem for 2.4 Ghz WiFi is the crowded and unlicensed
spectrum.
As for wireless networking ... I anticipate that there will continue to
be a number of complimentary/competing technologies, e.g WiFi,
Bluetooth, ZigBee, ZWave and even generic 433 Mhz radio for some time to
come.
--
-O- cheers = /\ /\/ /) `/ =
--O -- http://www.geekscape.org --
OOO -- an...@geekscape.org -- http://twitter.com/geekscape --
Reply all
Reply to author
Forward
0 new messages