4 ms·
I'll play the same tune I played the last time this was discussed: what is the cost? If the cost is above $30, I'm not sure I'd be interested. If they can bring
by ajays 15y ago
I'll play the same tune I played the last time this was discussed: what is the cost? If the cost is above $30, I'm not sure I'd be interested. If they can bring the cost down to $10, that would be ideal.
More generally speaking: how difficult is it to build one of your own? I'd like to build a sensor which is powered by, say, 2 AA batteries, has wifi and a, say, accelerometer. What would it run?
- joezydeco 15y agoIt would be pretty cheap, except for the wifi part. Those modules seem to be running in the $50-$60 range for onesies, not including antenna (or making room on your PCB for a screened one)
- ajays 15y agoWell, every module doesn't have to have wifi. It could be Zigbee, with a base station(s) that interface with wifi.
- dkersten 15y agoTrue, though that gets back to your original point, since Zigbee tranceicesr are also (relatively) expensive. I am currently working on some wireless hardware myself and while my devices do not need to be connected to the internet, they certainly could be since they do need to connect to a computer. Currently I am using ~$15 bluetooth modules for devices that can cost more and won't have more than a few instances deployed in one place. I will probably use wifi for high-cost, many-instance devices and I am currently evaluating solutions for low-cost, many-instance (though currently I only need low bandwidth one way communication for these). So far I have had some success with cheap rf-link mopules such as these[1], but am investigating cheaper/smaller/more-full-featured alternatives[1]. My target for these devices is sub €10 (~$13), which includes cost of microcontroller, rf module, passive components, circuit board and packaging. My point is basically that sub-$10 may be possible depending on requirements - base station or standalone; one way or bidirectional transmission; point-to-point or broadcast; size requirements and so on. I think for some of the applications mentioned in the article, sub-$10 is very possible. Of course you lose the single general purpose device appeal, but perhaps a small number of more-specialised devices could be available and you can then mix and match, especially if everything works off a single base station. [1] http://www.sparkfun.com/products/10532 http://www.sparkfun.com/products/10532 they are ultra simple, but also quite bulky [2] 2.4Ghz RF transceivers from Nordic Semiconductor among others
- joezydeco 15y agoI think the way out of this problem is to skip Wifi, Zigbee, RFID, NFC, and all the other fancy protocols. The chips are expensive, layout and shielding is a bitch, and you don't need a 54-mbit link to send the current temperature once a minute. Maybe the way to go is to create the simplest low power RF-transmitting circuit possible that can be driven by an Arduino or other simple microcontroller and make a simple protocol that tries hard to avoid collisions.
- dkersten 15y agoAgreed. I doubt each individual sensor needs to transfer more than a few bytes each time and probably not more often than a few times a second (and thats the extreme case - like you said, once a minute is more likely). This can easily be done with a low cost, low power RF circuit (and low power would have some additional advantages over the design in the article - eg, smaller batteries = smaller device) and wifi connectivity can still be achieved through a hub/base station. You only really need one wifi connection per sensor installation. If you need special super sensors that have their own connection for some reason, then special case those, charge appropriately and leave the base sensors as cheap sub-$10 devices. A simple temperature or vibration sensor or whatever shouldn't need to cost much more than about $5 or so, nor would I be willing to pay all that much more than that (certainly no more than $12).