How to make my IR badge idea a reality

[Eric Ball]

The idea is similar to the DefCon 20 and the Build Brighton Light Brigade badges - badges which use IR to exchange information.

Each badge has an IR LED and IR receiver driven by a microcontroller powered by an onboard battery with a microSD card for storage. The idea is each badge is programmed (via IR) with the wearer's name etc. When two badges "see" each other they exchange this data which is then written to the microSD card (ideally with a timestamp from an onboard clock). If the badge also had a microphone, it could also record the conversation to the microSD card.

The first question is a reasonable estimate for what this (assume several hundred) would cost to have made?

 

[large_ghostman]

I would go the IRDA protocol route, its old but many micro's have it native, or simple RS232 I have tried with IR leds. I am not sure why but stationary I got good speed but soon as there was the slightest movement I had to go slow, so if it was me I would have 1 IR transmitter Led and 2-3 IR transistor receivers next to each other on each badge. I would also seriously drop voice and SD card if you are limited money wise. Depending on how many badges you may be able to use eeprom instead of sd card, this should save money and there is no reason you couldnt add rs232 for pc comms.
No idea of cost but how many badges we talking? and what matters most features or cost per badge?

 

[Eric Ball]

Thanks for the prompt reply.

The idea for the microSD was to make it easy for the wearer to get the access list of "who did I talk to at this event" (and also leverage the list for gamification). But it might be just as good to have a "readout PC" which would pull the data from the badge and send it via email / sumbit it to the game.

The target is ~500 badges.

 

[large_ghostman]

Thanks for the prompt reply.

The idea for the microSD was to make it easy for the wearer to get the access list of "who did I talk to at this event" (and also leverage the list for gamification). But it might be just as good to have a "readout PC" which would pull the data from the badge and send it via email / sumbit it to the game.

The target is ~500 badges.

Explain this bit for me please

(and also leverage the list for gamification)

It would help if you fully explain how you see this working and what its for. It might be easier/cheaper to use usb
Alot depends on this gamifacation, will the badges be read at an event only or read from the person's pc later?
If its only to be read at the event then a simple IR rs232 interface to a pc would be cheap, if you want the user to read the badge at home then the usb method has merit,with SD you have the cost of the card and the socket, with usb you have the cost of a micro usb socket (cheap)
If its just at the venue then its a simple little box type solution (cheap/easy)
It might help if you can answer some of the following

Rough idea at how much you would like the badge to cost (this tells how realistic you are)

Rough idea what you are willing to spend per badge (this will make you think realistically )

How many times are they too be used? is this a one off event or are the badges likely to be used again?

What kind of battery life you after? how long is the event? This matters as it narrows some options

How big are you thinking of making them? do you need big or small? Is it better from a game point that they are hidden or better they are some kind of prop as well and fairly large? Size matters

When are you hoping to have them by? Again useful in seeing if your a realist or dreamer

Are you wanting to make them or have them made?

Does it matter to you what chip manufacturer is used? If your making them do you have programing experience? If so name the chip and language please.

FILL IN YOUR BLOODY LOCATION that should be obvious

what range are you wanting?

Do you want/need user input? would it be better or worse if the person wearing the badge had to do something in order to activate it? (because I have idea's)

Are you paying alone or do the people using the badges pay?

Do the people wearing them get to keep them? We need to know if the information needs to be fixed or flexible, if more than one person will wear each badge then it may make a difference

Explain the game side of things to me, I think if I understand what they are used for I will have a better idea of what you need.
sorry for the questions but it helps to fully understand what is needed.

 

[Eric Ball]

Explain the game side of things to me, I think if I understand what they are used for I will have a better idea of what you need.

The gamification is things like "who has met with the most people" or "who has met with the most people from different locations". The goal is to encourage conference attendees to mingle rather than sticking with their "home" groups.

It would help if you fully explain how you see this working and what its for. It might be easier/cheaper to use usb Alot depends on this gamifacation, will the badges be read at an event only or read from the person's pc later?
If its only to be read at the event then a simple IR rs232 interface to a pc would be cheap, if you want the user to read the badge at home then the usb method has merit,with SD you have the cost of the card and the socket, with usb you have the cost of a micro usb socket (cheap)
If its just at the venue then its a simple little box type solution (cheap/easy)

The idea is the badge is unobtrusive and the process is automatic. People meet and their badges say "Hello, may name is" and that gets recorded automatically and then later can be used for the gamification and so the wearer has a list of the people they met.

It might help if you can answer some of the following

Great questions!

Rough idea at how much you would like the badge to cost (this tells how realistic you are) Rough idea what you are willing to spend per badge (this will make you think realistically )

The Trippy RGB kit (on which the Build Brighton Light Brigade badge was based) is $10 each - so that's a good starting point for IR + microcontroller + PCB & glue. So $5K for the 500 badges + $5K in one-time costs. Probably more than my boss would like for "neat to have", but not outside the realm of possbility.

How many times are they too be used? is this a one off event or are the badges likely to be used again?

At $10K, my boss would probably hope they could be returned & re-used.

What kind of battery life you after? how long is the event? This matters as it narrows some options

Good point. The whole conference is over several days. Does that put the project from coin battery to car battery?

How big are you thinking of making them? do you need big or small? Is it better from a game point that they are hidden or better they are some kind of prop as well and fairly large? Size matters

I'm thinking something about the size of a normal "Hello my name is" sticker badge. Again the Trippy RGB kit & BBLB badge aren't much bigger than the coin battery they use.

When are you hoping to have them by? Again useful in seeing if your a realist or dreamer

The next conference isn't until next year.

Are you wanting to make them or have them made?

Soldering 500 badges myself doesn't sound like that much fun - especially if anything is surface mount. Any suggestions on how to find a shop?

Does it matter to you what chip manufacturer is used? If your making them do you have programing experience? If so name the chip and language please.

For microcontrollers I've only played with the Parallax Propeller 1 (SPIN and PASM), but I've coded in ASM on the 6502, 6809, 8086, 68K, 56K and a few others I've forgotten. I can also code in C and can muddle through with almost any other procedural language.

FILL IN YOUR BLOODY LOCATION that should be obvious

I'm in Markham Ontario, the conference is in SF, CA

what range are you wanting?

I figure 2-3' - talking distance.

Do you want/need user input? would it be better or worse if the person wearing the badge had to do something in order to activate it? (because I have idea's)

My idea is it's automatic. I'm thinking all I/O is via IR - including the initial setup with the wearer's name and the read-out (since discarding the microSD is probably a good idea). The programming & read-out could be handled by a PC with appropriate hardware.

Are you paying alone or do the people using the badges pay?

Company / conference provided.

Do the people wearing them get to keep them? We need to know if the information needs to be fixed or flexible, if more than one person will wear each badge then it may make a difference

Probably best if the wearer returns them at the end. Then they could be reused for a future event. Again, I'm thinking the badge could recognize a "set wearer name" IR command.

sorry for the questions but it helps to fully understand what is needed.

Questions are great! They define the problem space and help you help me - muchly appreciated BTW.

 

[large_ghostman]

Ok a few more Question's

How do you want the user informed of the data?
By this I mean do you want them to know each night or at the end? do you want them to be able to download the data to there laptop or will you just get them to pass a info point on say the way out and have the data 'downloaded' then you email them the data whenever? This is pretty important and to me is one the things that would define what direction I would go in.

Some thoughts

I misunderstood gamifactation, I had images of some kind of role play game thing not a conference. This does make a difference

The size thing I have no idea what size your talking about as I dont know the products you mentioned, a link would be helpful but my initial reaction is these would probably need to be slightly bigger than you think but not much and certainly not abnormal/uncomfortable to wear.

The price to me puts you outside a normal shop but well within the realms of a good amateur.
5k for parts is going to limit some options, so forget recording voice and bluetooth (would have been ace), it still leaves some form of radio (just) but radio is complex as your basically looking at some kind of mesh network (IMHO).

BY THE WAY
I was expecting others to jump in by now but thats ok, I have to tell you I am 14 in July so no long in the tooth seen it all engineer, on the other hand I have a active mind not messed up with a lifetime of experience. So from this point on you have two choices
Cut the conversation and wait for better to come along

Stick with it if only to get some idea's
Would I make these?? I would be tempted to have a shot at a design and maybe get a prototype made up if I thought I could A) Do It
B) Prove I could Do It in enough time to give you options if I couldnt

Do I think I can do it??
I dont know but I am sure given a few days I could say one way or other as knocking up a demo badge should be easy enough and not cost me to do.

IR worries me a little in as far as it WILL miss the odd contact. You have a small problem with IR, if you make it short range then you risk missed connection's, but if you make it too far ranging then other badges will pick up the data when they shouldn't. You need to consider which case is better

Battery life depends, I wouldnt rule out something like a 2 coin cell option for 3 days of life, but again depends on how you want the data given out to the badge owner.

Making them reusable is easy enough, it just means one path instead of another.
I need to sleep on this for the night and have a think, there is alot of questions floating in my head but I need to think it through.
Its a great project, tomorrow I will be back with more questions and then should be able to give some thoughts, hopefully after that others will chip in and say yes I am being reasonable or no I am being an idiot. IR comms interest me and I have done stuff with IR before. I need to get a good list of question's in my head as it will dictate the option's you have, I certainly think its doable but would want to make sure first.
Also have a good think of what kind of information you want exchanged, I have some idea's but want your list first to see if I am the same wavelengh, Also be careful that you dont miss chances, this could be a really powerful tool if you think about it
a micro and built in clock could tell you alot about how well the conference went and if any changes you make have a positive or negative effect. The time each person connects to the other for and how often over the course of a day would let you plot movement on a room plan in something like matlab. In theory you could actually collect some powerful statistical information that would give you priceless feedback you couldnt get any other way. I will detail this alot more tomorrow after my question's. But I do see huge potential for a system like this

 

[4pyros]

You should be able to make it so you put it in a pod at night and it recharges itself and dumps its data each night.

 

[large_ghostman]

You should be able to make it so you put it in a pod at night and it recharges itself and dumps its data each night.

I cant see that for that sort of money, plus you have to get delegates to do it, the data could take all night to get. I am sure I can get a low energy micro running 4 day no problem, my recent foray into energy harvesting has taught me alot on making battery stuff last. There are bigger problems as I see it, People normally wear a badge on the left so two people facing each other have badges not opposite, the other problem with cost is a $ in the states buys more than a $ in the uk, for me it would be design and have made in china. The problem with data collection could be solved by a simple panel placed by the door on the way out, it would collect the data from each badge, or a couple of panels in the room and one at the door, then every time people go near the panel it downloads and clears the badge.
I wouldnt use coin cells but there are other small batteries about that might work, some of the special camera batteries are pretty good, if I can get range with about 2mA then we are on, the rest is just very clever software. I have done some EFM 32 bit stuff from energy micro, they would cruise this on a coin cell and probably get you around 4 weeks, there is a good and fairly accurate battery adviser in the stuff, running it through there with some guesses says its a goer, the trick is not to wake the core until needed you can run a timer on a DMA channel and it sends a beacon this wakes another badge if present, that responds and wake's the sending badge to another level, it sends its ID and ask's for ID. Long story short if its the same badges talking that have been talking more than say 45-60 seconds then they just time stamp no need to recollect other data as the badge know's it. If its a new badge that it hasnt spoke to before OR if its a badge that there has been other contact with that day they exchange different data (depending how deep they want to analyze this data). " badges connecting for say 5-6 seconds dont exchange details, 5-6 seconds isnt really a meeting its a quick hello at best! So to me wouldnt count.
What interest's me and my OCD is the data you could collect and then map, that would show key people that hold others attention and attract people, these you would target for other conferences, the people that sit in a corner and dont talk to anyone you may not want again as they dont give much.
Alter the room layout and it would show if you can influence room flow and propagation/migration (again an interest as it fits my OCD). The main cost (forget battery thats going to have to be an extra) will be SMD IR transistor, I have some here that work well for other detection applications better than IR diode detectors). Someone good with stats could have a real field day, depending on what the company doing this does I would get hold of a uni prof for psychology I am betting they would share some the cost for the data set . Also they could make this a powerful tool

ANYWAY back later with questions

 

[Eric Ball]

Build Brighton Light Brigade badge: **broken link removed**
Trippy RGB kit: https://www.makershed.com/Trippy_RGB_Waves_Kit_p/mkce4.htm
DefCon 20 badge: https://www.wired.com/2012/07/defcon20-badge/?viewall=true (Parallax Propeller powered)

Note: although I have hands-on experience with the Parallax Propeller 1 - it's overkill for this project, thus I'm processor agnostic.

 

[large_ghostman]

Build Brighton Light Brigade badge: **broken link removed**
Trippy RGB kit: https://www.makershed.com/Trippy_RGB_Waves_Kit_p/mkce4.htm
DefCon 20 badge: https://www.wired.com/2012/07/defcon20-badge/?viewall=true (Parallax Propeller powered)

Note: although I have hands-on experience with the Parallax Propeller 1 - it's overkill for this project, thus I'm processor agnostic.

A few question's on some other stuff, what are you hoping these badges will do?

What does the company they are for do?
This is kind of the main question in some way's, for example do you set up conferences, or do you do something else but rely on conferences for some reason? What I am after is understanding what you could get from this.
You say the chip is overkill for what you want but there could be another reason for this, depending on the importance of the conference to your company you could miss a very good chance. Say you do the conferences to launch new products or connect with new customers, then someone clever could take the kind of information gathered and turn it into one hell of a powerful sales tool.
Sorry to be blunt (I am just that way) but I get one of two feelings about this and your perception of it
1) its just something we want to do for no particular reason and we do these conferences because its always be done that way.

2) You genuinely dont see the potential of the kind of information these could gather for you?

So to get to the bottom of it explain why you do these conferences adn what you hope they do for the company, if you dont organize conferences as a company then what DOES the company do/make?
500 people is alot in a room, if you use a conference to sell something or make contacts then how key people are received matters.
At the moment its just guess work but take the following situation.
Say you sell something and at these get together's you have some sales people. On the face of it the sales man with the most sales is the better salesman, but is he? A system like this would show who spend what time talking to whom and who they then spoke to, it could be you have a guy that talks to everyone for say 20 mins but sells nothing, yet he could be the man that sparked the interest.
Each time you change the room layout at a conference you should see a different pattern of movement between people, you could use this to channel people to key notice boards or key people. I thought of around 60 things like this that you could get from a badge system.
If its just a gimmick to get people to mix then your missing a great opportunity to influence how they move, to be honest if its a gimmick like that then just get cards printed, tell people the person who collects the most cards a day win's XYZ.
It would save you money and give you the basic data, because the trouble with making it into a game only instead of a game like tool is your going to have people focus on the wrong part, what you want is to encourage the right kind of mixing and not just a kind of Hi I am XYZ type of game where they then move on.
There are ways to overcome that though.
Thanks for the links I will take a look

 

[dr pepper]

Interesting project, shouldnt be hard to do, you can generate tone burst rs232 suitable for tssop infra red receiver modules easy enough with a couple of external comps and a pics uart and pwm block.
The only issue I'd see with that is bright sunlight could swamp the sensor, you'd have to try it and see.
You can get sd card socket/interfaces from china cheap.

I like your avatar, interesting parody.

 

[large_ghostman]

Oh Dear I looked at the links!

You have based some assumptions on wrong info, no idea what the first link is or what it does, but I assume it dosnt communicate to another badge, non do except maybe the last one might, but there you have this quote

"Clarke says producing the badges in the United States increased the cost, making them the most expensive badges DefCon has ever made. He won't say exactly how much they cost, but notes they're more expensive than badges last year that were made of titanium and cost $10 apiece."
Last badges of titanium and $10 the ones in the link cost more, so much more he wont say. You then count the parts and guess the cost of parts alone would make them $20 minimum without assembly or design fee's.

I still think what you want to do can be done, but there is no way any of those link give you anything to start from sorry.

So its going to boil down to what your companies does and what these badges could do for them, the $10 badges seem to be non communication ones, this makes a difference, as soon as you want badge to badge talking at $10 your in limited choice mode. IR is cheap but in some ways harder for this, to be honest If you could guarantee everyone had a smart phone I would get a app developed, JOB DONE

Anyway tell me about your company and what it does

 

[large_ghostman]

Interesting project, shouldnt be hard to do, you can generate tone burst rs232 suitable for tssop infra red receiver modules easy enough with a couple of external comps and a pics uart and pwm block.
The only issue I'd see with that is bright sunlight could swamp the sensor, you'd have to try it and see.

I like your avatar, interesting parody.

I am going to argue with you on this yipee!
I have you my friend on this
On the face of it yes simple, but read the posts again. then think about it , you have stray beams to consider and the fact that people opposite each other will have the badges not aligned. When I did the line follower robot I wanted IR comms from a handset to it, sounded easy ended up but finally

 

[large_ghostman]

Also you need to store and retrieve information

 

[dr pepper]

Yes your right, you only really see issues when you start a project.

There are solutions to most problems but that involves time and complexity.

More leds will improve the angle comms is possible at.

You'll also if there are many of these devices need to devise a protocol, maybe something like dallas's one wore protocol, so that all devices can communicate without contention.

Have you seen those little modules on ebay for using sd cards, just the job for this.

Constructive arguments allways accepted.

 

[large_ghostman]

Yes your right, you only really see issues when you start a project.

There are solutions to most problems but that involves time and complexity.

More leds will improve the angle comms is possible at.

You'll also if there are many of these devices need to devise a protocol, maybe something like dallas's one wore protocol, so that all devices can communicate without contention.

Have you seen those little modules on ebay for using sd cards, just the job for this.

Constructive arguments allways accepted.

Yes I have seen those modules but you need to keep cost in mind, $10 per badge isnt much, especially when you say stuff like more Leds will improve comms, yes but every time you add a Led you you add cost, I think $5K for doing a job like this is ok for a starter job, but you have to convert into £ to understand. $10 after exchange rate and fee is what around £3,700 so lets be safe and say £3,500 , thats £7 a badge! for parts, 500 is a awkward number, not enough for high volume prices but enough to make hand build a pain!
Also you need a blank bank to start with and a pcb made, then add a micro and already your making a dent in the budget. Add a battery!!! Well dont add a battery I think £7 is doable but without the battery, it depends though.
I am thinking in some ways its a gut feeling job, in order to really get an idea of cost you would have to play around a bit, that takes time and add's cost, no one is going to play around with code etc without getting the job first.
£3,500 for the kind of hours needed is ok for joe blogs wanting to start out, I would jump at it via dad of course if I was sure I could deliver on time on budget and it was kickass quality.
But I wouldnt do that until I had enough information to make a good judgment, I think the key is a simple serial protocol with low power (2-3mA) IR leds.
What really matters is what we havnt been told yet. If its just a bit of a gimmick then you can do the same thing cheaper without all the electronics, but if you could get really useful data from it then I think there are other options regarding pricing. I have much going around inside my head but first I want the other stuff answered.
I have thought of a software map for all this and it wouldnt be too hard, but would need to be clever and not just a constant beam the same message type thing, that would kill the battery quickly.

 

[dr pepper]

A very professional answer lg, quite right.
Your also good at diplomacy.

 

[large_ghostman]

Me and diplomacy are rarely said in the same sentence!!! Thanks .

 

[dr pepper]

Maybe your having a good day, keep it up.

 

[4pyros]

The only issue I'd see with that is bright sunlight could swamp the sensor, you'd have to try it and see.

Just a thought; I don't think you will see much daylight in a convention hall.