Slashdot Mirror
Wireless Audience Response Systems?
kjeldahl asks: "I've got some project ideas involving what I have learned are named audience response systems. These are small devices (typically a small keyboard with a few buttons) that can be handed out to an audience where they can participate in some kind of 'voting' or selection process where they press some buttons and their selections are recorded by a central server. Looking further, the wireless options include RF and IR, although IR senders typically conflict with eachother in environments with lots of people pushing buttons simultaneously. To be useful, one typically needs to be able to know the various devices apart (device X sent keypress Y). Some of the more advanced devices even include two-way communication with display abilities. Anyway, these devices seem very specialized and kind of tied to the platform (including hardware and software) - in effect quite expensive. Does anyone know of open audience response systems which use well-known technologies which can be adapted for use for audience response systems? I would guess the candidates include wireless keyboards (assuming they do not conflict with eachother), Bluetooth and WLAN-based systems."
It's called "applause". If you don't hear any, then the show sucks...
"Freedom means freedom for everybody" -- Dick Cheney
How come you don't make the keyboards embedded into the chair in front of them, or pullout keyboards like some college lecture rooms have pullout 'deskspace'?
This is rather simple and conversion from ps/2 or usb to cat5 cable for the sole purpose of keyboard data isn't very difficult.
It seems to me that a large portion of the inherent costs of a system like this is the hardware. What are your per-audience-member costs for a typical system? Maybe a bunch of wireless palms and a smallish pc server could do the deed?
meh.
I call it "public entertainment-induced manual impact resonance" or "PEIMIR".
The PEIMIR system is based on the simple fact that striking two surfaces together makes an audible mechanical wave (a "sound", if you will).
Now, what use is this you say? Well, it turns out that the human body itself has two surfaces useful for this purpose: the palms of the hands!
With minor training at the beginning of your show or talk, you can teach your participants this simple technique. At first they may not be willing, but I've found running a lottery with a small prize of $5-$10 is sufficient inducement.
Then throughout your talk, you can use the relative intensity of the PEIMER mechanical waves (the "volume" if you will, though that term is abiguous) to gauge the success of your talk.
here's a tip I've learned in my "PEIMER adventures": sometimes high-intensity PEIMER waves don't always mean success: for instance if you here NONE throughout your talk, and then you hear it with high-intensity at the conclusion of your talk, that means they are pleased you are leaving.. so your talk was bad! It took me a while to figure this out but it was quite an eye-opener when I did!
Good luck! My PEIMER system is patent-pending but I do allow non-commercial use for free! Contact me for details at my slash dot address.
A. Coward
PEIMER facilitation specialist
the PEIMER foundation
Previewing audiences are frequently handed little boxes with a dial on it which they use to dial in their approval of the show. If they like it a lot they can dial all the way up. If they hate it they can dial it all the way down. If they find something mildly amusing they can dial it to somewhere in the middle.
It's instantaneous and doesn't have the problem of a hundred people pounding on a keyboard in the middle of a show.
If you use a very wide bandwidth to the signal, then you can have wireless transmitters in the studio... but why is it now that we must take a method that works and change it so? I mean we all remember those dials they used in America's Funniest home videos. Also, for Candian Idol, they let people vote via text messaging on their mobile phones. That worked well and they got better voter turn out than our federal elections. Cheers.
while(1) { fork(); };
bluetooth!
ok, kidding. sort of. You could use PDAs (ok, not cheap, but lots of people have them) and use a 'base station' for each clump of people.
I wouldn't want to implement it (any idea what the collision issues would be, anyone?). but hey, it's an idea.
how about cell phones for that matter. or sms-enabled pagers. seems like there are devices out there that do similar things, just not in such enclosed spaces.
I think that Chuck Wollery could give you some advice.
Maybe I'm missing something, but if you hand out little wireless enabled gadgets to the audience that are not securely tethered to the location where they'll be used, aren't you asking for them to be stolen?
Obligatory: anyone want to get ahold of these and get ahold of a free wireless device?
Simply shrink each and every audience member to be the size of a nickel, then have up to 104 of them stand on your keyboard. If they like something, have them jump, thus depressing their key.
The typical audience response system involves mere dozens, or a couple of hundred at best, user stations, and for TV live broadcasts or tapings they are typically wired (so as not to interfere with audio or video, esp. wireless mics). Any wireless system of sufficient numerosity to overcome the efficiency of merely using a wired system would be subject to packet collisions and crosstalk (infrared, from studio lighting; RF, to/from various other RF receiving or emitting devices). Real TV shoots burn too much money per minute to incur the risk response-system debugging delays.
Audience response systems (wired, for maybe 10-20 people, max) are used in TV show test-audience previews, e.g. at the Venetian or MGM Grand hotels in Las Vegas where you can earn $10 by previewing the latest NBC or CBS programs, respectively. Tell the nice lady you are between 18 and 49.
(MAD magazine once theorized that new network TV shows are selected according to how many preview-audience members fall asleep on their "I Like" buttons as opposed to the number of testers falling asleep on their "Don't Like" buttons.)
The best known and most widely used audience response system in television is the Applause light where a wired system connected to a single SPST switch illuminates an Applause sign. For some reason, this device, when activated, causes the audience to respond with applause, whether or not the subject matter is praiseworthy or even entertaining (e.g. the appearance of a new washing machine on The Price Is Right).
At Rutgers University's Physics Lecture Hall there is a PRS (Personal Response System) setup. Basically, every student who ever takes a Physics class has to buy a $30 transmitter to participate in class. It works via IR so the system can get overloaded with a large class. However, the IDs of the units responding show up in a dynamic list on the projected screen as you respond (so you can see if your answer got registered). The system then displays a graph (or presumably other information) detailing which answers were chosen. It's good because it can't hurt your grade if you are there to press the button (you get one point for pressing the button, two for being right, out of a possible one point). It's bad because you lose points off your PRS portion of the grade if you don't go to class or forget your transmitter. You cannot borrow someone else's because an ID is hard coded into the device and then one to one associated with your student ID.
"Nature doesn't care how smart you are. You can still be wrong." - Richard Feynman
I've only had experience with the H-ITT device (http://www.h-itt.com). It's IR and has lots of problems when everyone uses it at once. Based on my experience with this, I think some kind of rf-thing is better. Perhaps some like of 802.11 or something that can handle overlapping requests.
For any that are interested, I've reverse-engineered the serial line protocol for H-ITT, which would allow one to write their own clients for the hardware. H-ITT only gives out the software for Windows, and dispite the academic demand refuse to develop for Macintosh, so we are working on a Mac app for our faculty.
If everyone has a wireless capable laptop they can bring in, just write up a webpage with Java servlets or something, wire the room with wireless access points and servers, and you're done.
If not, that previous wired keyboard option was a good idea, but does anyone know how much it would cost to build a wired keyboard to wireless converter? If you're only going to rely on a few buttons, I don't imagine that it would be a lot..
As a teacher I was given the opportunity to use a system my district purchased called classroom performance system. Basically it's a bunch of small infrared "remotes" with buttons A-H on them and an infrared receiver. It's designed for classrooms, so it's oriented around a question-response mode and assessment, but it is also very useful for rough surveys, votes, etc. In particular, I am told that the state of South Carolina (luddites that they are) used this system to enable an enormous gathering of teachers to vote on which questions should or should not be included in their high-stakes NCLB-compliant test, the PACT.
Dustin
Disclaimer: I don't work for eInstruction; I just used their product when I was teaching, and was very impressed.
Please refrain from tasting the knob.
Never fight naked, unless you're in prison...
I've built a system like this before, and I can tell you the one reason why they don't use wireless handheld gadgets. People will STEAL the damn things if they're not wired down. This is another classic case of why engineers should not be put in charge of product design. They're so caught up in the coolness of some new technology that they forget that PEOPLE are involved.
To me, this sounds like the perfect application for a short-range wireless protocol such as Bluetooth. With the integrated Bluetooth chipsets coming out these days, manufacturing would be pretty inexpensive; I mean, it only has to read a few buttons.
If you connect every single device to a central node, that's a LOT of bandwidth you're asking for. If you have every device capable of gathering information from its neighbors and passing it along, you can get the information to aggregate itself along the edges of the audience, where it can be collected. The devices could deduce where the edges are by counting the number of nearby devices.
Here's a single-chip solution: the Zeevo, a microcontroller and Bluetooth chipset rolled into one. Now, Bluetooth development kits are HORRIBLY expensive, the page I linked contains a kit selling for about $8,600! But this product looks VERY interesting, it is a Bluetooth-to-serial development board, but you do get control over many Bluetooth parameters and it mentions firmware download. The cost is only $75 (or Euros to be exact). I think that would be an excellent way to get a hackable Bluetooth platform to play with, depending on how much of the firmware you can reprogram.
...
Look, I don't want to seem impatient here; but, could we just pick one wireless system now, so we can get started hacking it right away?
Sometimes I worry that I'll develop Alzheimer's disease, but no one will notice.
Since the bandwidth needs are almost nil, it seems like there ought to be some really sleazy, simple, and effective way to do this with RF. Perhaps have a central device that polls each voter sequentially for their answer(s) several times a second. This would get rid of the need to sense collisions or other stuff that happens when multiple conversations take place simultaneously. Good luck!
lock in your votes!
"Sic Semper Tyrannosaurus Rex."
Our University used to have one of these systems that they paid thousands for... Then some people in our physics department figured they could make the same thing for about 1/10 the price. They designed an RF transmitter and reciever and wrote software (works via serial port, connects through cat5 cable, and recievers can be setup around the room daisy chained to eachother).
The remotes you pass out to the audience (students in our case) are the size of an ink pen. They are very inexpensive and the recievers are inexpensive as well. The drivers for the recievers and the polling software is all downloadable from their website...
have a look: H-ITT
The recievers are around 80-100 dollars, and the remotes are very inexpensive (i believe our bookstore sells them for like $15 or $20)
Two infinite things: your stupidity and mine. But I'm not sure about the latter. If my sig offends you, I'm sorry.
I think the poster of the question needs to let us know why wireless over wired? We're assuming the audience is sitting in seats, not moving. So why do you need devices that move freely, are able to be stolen, and have all the problems you mentioned?
Trolls lurk everywhere. Mod them down.
You can use a bunch of old PDAs that can use BlueTooth modules. Thats if you got lots of them...
Use RFID tags. Create a box that masks different tags according to what buttons are pressed, and have receivers around the room picking up the different tag IDs.
With this system, you could 'transmit' binary numbers from 0 to 2^(number of RFID tags), as each tag would be either 'on' or 'off' (unmasked or masked.)
You are laughing... but this might be a workable solution.
Don't shrink the audience... expand the keyboard!
Get an industrial-strength keyboard, and dissect it. Thoroughly. Read the previous slashdot post from the guy who took his keyboard and an old typewriter, and made the typewriter act *as* the keyboard.
Then give each member of the audience a "key" (whammy-happy-button) to press when fun things happen. Have a few PCs that "read" the keys pressed from the keyboard, and do a frequency analysis comapred to events that happened during the show. Depending upon how fast/repetetively they hit the key and how fast the kayboard could register keystrokes, you could be in business.
You could track each individual user (PC#3, 'f')'s responses, and the set in aggregate.
With respect to hardware, you would need enough PCs to cover the audience / (max freq read), iven that there were enough keys per keyboard. have extra / sufficient $ and time? Give them 2-3 keys (happy, sad, funny, etc.) and do more in depth statistical analysis.
The one item that you would have to "record" would be the timing... some sort of an FFT (fast fourier analysis) would most likely be appopriate, depending on how you captured the information. You would need to capture/indicate the times when they *weren't* pressing the button--this is important as the times they were!
Good luck!