A turn-key home automation RFID add-on to HomeSeer and Crestron using Wavetrend hardware and customized software has been available for about 15 months. The Homeseer package was _almost_ reasonably priced when they had a beta-bundle-bargain when it first rolled out.
I'd be interested in learning about experience with RFID for HA in general and especially less expensive long-range, homebrew alternatives and other than what is discussed in the Homeseer forum
An article in the Marc 30 2006 hard-copy edition of Electronic Design describes G2 Microsystem's
new active RFID chip.
The $12 (in quantity; scheduled to ship 3rd quarter 2006) G2501 chip performs active RFD over standard 802.11b WIFI. Battery life using two AA's is projected at 5 years.
The chip is programmed through a url and "uses existing WIFI wireless access points on the network, meaning no new network installation is required" if a WIFI network exists.
Spatial location of the tag to within 15 feet can be accomplished by triangulation using three WIFI access points, or to within about six feet using ISO 24370/ANSI 371.1 Time Delay of Arrival (TDOA).
The chip also has two standard passive RFD technologies built in: Electronic Product Code (EPC) and a standard 125mhz magnetic. These are powered internally, don't need the WIFI transceiver to turn on and "can talk to other existing RFD systems in their neighborhood".
GPIO, UART, SPI, current loop and ADC interfaces are built-in and on-chip. The latter are specifically designed for sensor interfaces such as temperature, humidity, light, radiation and so on.
| This strikes me as a terrific idea. The whole reason I'm interested in RFID | is to know if my cars are in the driveway. If both are gone, temperature | adjusts, doors lock, roomba vacuums... etc. | | So I don't even need a small battery powered chip. It could be bigger. A MCU | with an off-the-shelf USB access point. Or one of the smaller chip | solutions, but not wallet-sized. | | Anyone have any ideas?
I've gone through several generations of this. The first uses a moderate sized old laptop (Toshiba T1910) with a PCMCIA 802.11b card in each car. The second used a smaller computer. The third uses a modified Aironet BSE in a WGB-like mode. For the first two generations I run my own tcp/ip stack tweaked to reset the ICMP timestamp response value to 0 on startup. The third generation provides an uptime value via SNMP. First and third generation are currently in use.
I power the 802.11 device (computer or BSE) only when the engine is running. A daemon running on my automation computer periodically queries each car for its uptime and generates events when a car appears or disappears. It infers from the uptime whether the car is starting up or coming into range (for appearances) or shutting down or going out of range (for disappearances). The inference based on uptime was intended to be a temporary hack (to be replaced with either explicit status messages or signal strength tracking) but it has worked so well that I've been reluctant to introduce more complexity.
It is critical for the car-based device to boot up and start responding to queries before the car gets out of range when starting/leaving. Even many DOS-based laptops waste a lot of time in the POST, making them useless. The Aironet BSE platform (same hardware as WGB) runs the code from flash so it doesn't waste any time decompressing code and such. None of the other Aironet platforms are suitable. I haven't investigated other brands, mainly because the Aironet platforms use the Aironet PCMCIA cards and they have great range...
Thanks for your response. I've been mulling this over for a few weeks now, trying to figure how to slim this down even more. I was thinking of something as simple as a USB access point in the car, with a script to detect when its mac address showed up on the network. Problem there is one you've already identified: You need some sort of uptime to tell the difference between an arriving car and a starting car.
This is very interesting to me, and I'll continue to persue it and let you all know what I come up with.
Anyone else have any ideas? Wifi instead of RFID.. the $500 cost savings have me drooling.
E. Lee Dickinson Entertainment Design and Technology
| Thanks for your response. I've been mulling this over for a few weeks now, | trying to figure how to slim this down even more. I was thinking of | something as simple as a USB access point in the car, with a script to | detect when its mac address showed up on the network.
I'm not sure why you'd want to involve USB, and you don't want an access point so much as a client device.
|Problem there is one | you've already identified: You need some sort of uptime to tell the | difference between an arriving car and a starting car.
Happily, the uptime counter is available from just about any SNMP agent. As I mentioned, the Aironet BSE device (e.g., BSE340, BSE342) is a good choice for the in-car device. Sometimes you can get them quite cheaply on eBay. Even if you don't use Aironet products for your wireless infrastructure the BSE will still be able to connect and you will be able to query its uptime. (You won't be able to use it as a bridge to an in-car Ethernet, but that may not matter.)
| The price of WIFI routers has dropped below $50. | | Linksys open sourced the WRT54G Linux firmware in July 2003. | |
| | They boot in a few seconds.
Do you happen to know more specifically how long it takes from power up to association complete? (The Aironet BSE34x takes around 20 seconds which by weird coincidence is also how long the T1910 takes using DOS and my stack.)
| All one would actually need is a 555 IC (or other timer) that would wait x | minutes to power up any detectable characteristic of the router to | distinguish between coming and going.
Doesn't the WRT54G have an SNMP agent with an uptime counter?
From the time I replug the power to the time I can regenerate the login screen is about 9 seconds. I presume this means that it has finished booting but don't know that for certain. This is with version 2.04.4 Linksys firmware. There are various hardware versions of the WRT54G that apparently are best distinguished by looking at the board. I have three WRT54G's different vintages. I have one with fried ports that I could open up.
Dunno about any uptime counter. Wouldn't surprise me that this would be a feature added by one of the several (many?) alternative firmwares.
Interesting, Cisco/Linksys now sells the WRT54GL version that is specifically meant for the Linux community (about $70 mail order). They were able to get their firmware on less flash, so the currently sold 'normal' WRT54G won't load the previous firmware. Hence the -GL (L for Linux) version. Nice touch on the part of Cisco.
If this is something you'd like to pursue, I have two WRT54G's that I can send you. One works perfectly, the other has at least one blown ethernet LAN switch port (lighting damage ABIK) but if experience serves, the RF and WAN side may be fine. I may still have your snail mail from another project. I've been procrastinating on replacing the Linksys and re-installing the hardware firewalls and dual WAN router that were removed in a recent infrastructure upheaval and this would be the impetus to do so.
Hi Dan, I can only give you my opinion on the Linksys firmware because I don't remember seeing any mention of it (seems to be typical of low end routers). The 3rd party firware has support for SNMP (and not just uptime). I'm currently running OpenWRT on my WRT54G, WRT54GL and Netgear WGT634U. All have SNMP support with OpenWRT.