[00:00.000 --> 00:05.140]  As Greg said, this is meant to introduce the attendee metadata system.
[00:05.900 --> 00:09.480]  Up here on the stage are some of the people involved in the project.
[00:09.740 --> 00:16.320]  We actually had about over 20 people at different times involved in getting it running, which it will be soon.
[00:18.460 --> 00:19.520]  This is Matt.
[00:19.720 --> 00:22.300]  He's been one of our co-infrastructure leads.
[00:22.520 --> 00:27.200]  Infrastructure is running the cables everywhere and the readers for the RFID system.
[00:28.020 --> 00:35.900]  Milos is from the CCC and he has openbeacon.org and he developed the RFID system that we're using.
[00:36.300 --> 00:38.120]  So he's very smart.
[00:40.220 --> 00:43.620]  And then on the end is one of our coders, Neo Amsterdam.
[00:44.220 --> 00:50.200]  And Christina is coming back in a minute with her laptop so we can show you the website.
[00:53.960 --> 00:54.760]  Let's see.
[00:55.100 --> 01:04.120]  The basics is that we're using an online survey and matching the survey data with position tracking RFID.
[01:04.800 --> 01:15.080]  And we're combining that data in the same database to figure out what people are interested in, what they're doing at the conference for about half the people here.
[01:15.080 --> 01:19.020]  We have 1,400 of those little RFID badges floating around.
[01:19.800 --> 01:25.120]  And we were kind of expecting somewhere south of 3,000 people to show up at the conference.
[01:25.400 --> 01:28.120]  So roughly half the people showing up should be able to get one.
[01:30.520 --> 01:36.620]  So I'll start off on the why part of why this was started.
[01:37.100 --> 01:45.560]  About a year and a half ago, I was on this new forum that they put up on hope.net, which is talk.hope.net.
[01:45.700 --> 01:48.640]  It's a little forum board where people are talking about HOPE.
[01:48.640 --> 01:52.800]  And they put a section up that said new projects for The Next HOPE.
[01:53.900 --> 01:58.380]  And I'd been to H2K2 and The Fifth HOPE and HOPE Number Six.
[01:58.960 --> 02:05.860]  And it occurred to me that we get in these big seminar rooms and we have all these people and one person is standing up here talking.
[02:06.340 --> 02:11.480]  And there are all these people out in the audience who are interested in the thing they're there to hear about.
[02:11.480 --> 02:19.380]  So getting them to connect to each other was kind of something I was interested in.
[02:20.080 --> 02:28.000]  And on January 17, 2007, I posted this message titled Attendee Metadata.
[02:28.760 --> 02:33.980]  I'm not a programmer, which causes me no end of frustration with the number of projects that spin out of my head.
[02:34.060 --> 02:37.940]  But I have this idea for getting people with similar interests to meet during the conference.
[02:37.940 --> 02:48.080]  The basic idea is that we have a web form that collects information about your interests and uses that information to help you find like-minded HOPE-goers.
[02:50.020 --> 02:55.780]  I've been talking about how hospitals were using similar systems to get patient information.
[02:56.460 --> 03:05.440]  Instead of having you sit there with a clipboard, they're using web terminals to gather that sort of information to get the doctors more information in advance.
[03:05.440 --> 03:11.960]  And in this case, we're helping each other get more information about each other's interests in advance.
[03:13.980 --> 03:23.940]  And I also said that this could compare to the Myers-Briggs questionnaires that kind of help you figure out what your career path is or might be.
[03:24.880 --> 03:35.220]  So what I said was that I proposed that we set aside an hour during the conference where a bunch of tables would be set up with different symbols.
[03:36.020 --> 03:39.960]  And at each one, there'd be a big bottle of Mountain Dew and a bunch of pretzels.
[03:40.720 --> 03:54.140]  And we'd have people go through the website and figure out what their interests were and the computer would spit back at some point a symbol that they would use to find other people with that interest at the conference.
[03:54.140 --> 04:08.380]  And so after I posted that message, suddenly people were suggesting physical tags that people would actually wear so they could connect to each other without going to the tables sort of thing.
[04:08.640 --> 04:13.760]  And then Thurman suggested that we start using smart cards or RFID.
[04:15.020 --> 04:18.980]  And then people were kind of spinning around.
[04:19.120 --> 04:24.320]  Emmanuel came in on the forum and said, this is an idea we've been talking about for a while and you'd love to do it.
[04:25.100 --> 04:31.460]  And it went in circles for about six pages and died September of last year.
[04:33.700 --> 04:37.360]  So then the ideas were still kicking around in my head.
[04:37.460 --> 04:41.020]  And so back in March, I was directing an off-Broadway play.
[04:41.020 --> 04:45.200]  And Kupo was playing a part in the play.
[04:45.680 --> 04:51.140]  And Ray Trigger and Steggs came down to see the play.
[04:51.400 --> 04:53.200]  And afterwards, we went up for food.
[04:53.200 --> 04:59.540]  And I kind of pitched them this idea about the attendee metadata and how we might put it together.
[05:00.860 --> 05:08.840]  Because I think we were talking about the build-a-printer project that Druid was doing, where they were trying to build a giant printer that prints the side of a building.
[05:08.840 --> 05:11.340]  I don't know what the status is on that project.
[05:11.560 --> 05:18.160]  But it seemed like a great idea to get people together to do a cool little technical project that you could show at the conference.
[05:18.420 --> 05:20.740]  And so that's kind of where this got started.
[05:20.840 --> 05:23.100]  And a week later was my birthday party.
[05:23.100 --> 05:25.820]  And we got like 20 more people interested.
[05:25.880 --> 05:28.080]  And it just kind of snowballed.
[05:29.100 --> 05:32.340]  So then the call for papers came out for HOPE.
[05:34.120 --> 05:38.040]  And I submitted a proposal to talk about the project we were doing.
[05:38.200 --> 05:40.280]  And the email came back.
[05:40.500 --> 05:42.620]  Are there two talks about RFID?
[05:43.280 --> 05:46.040]  Or rather, are there two RFID projects?
[05:46.040 --> 05:56.000]  And so it turned out that Emmanuel had been wanting to get the CCC system that Milos had developed over to HOPE.
[05:56.200 --> 05:58.800]  And so we started looking at that system.
[05:59.040 --> 06:04.400]  And Aesthetics and Openfly and I went and met with Emmanuel and Bernie at Notacon.
[06:05.020 --> 06:06.680]  And talked about the system.
[06:06.880 --> 06:10.040]  And we were just kind of sitting around a big table throwing ideas back and forth.
[06:10.040 --> 06:14.220]  And right there Emmanuel put about 12 grand down to buy the RFID tags.
[06:14.440 --> 06:18.140]  And so that's how the project kind of got launched.
[06:18.580 --> 06:23.940]  We had crunch numbers about the Motorola 915 megahertz system.
[06:24.340 --> 06:27.740]  That it's got a range I think of about half a meter.
[06:27.940 --> 06:28.960]  A little more than half a meter.
[06:29.700 --> 06:32.520]  So we were going to put little boxes at the doors.
[06:32.760 --> 06:35.270]  But Milos' system is about 50 meters.
[06:35.760 --> 06:36.060]  Is that right?
[06:36.300 --> 06:36.480]  Yeah.
[06:36.640 --> 06:36.820]  Yeah.
[06:37.220 --> 06:38.600]  So much longer range.
[06:38.600 --> 06:41.380]  But the Motorola system was also passive.
[06:42.220 --> 06:43.660]  So our system is active.
[06:43.800 --> 06:45.680]  It's got a battery and actually sends out a signal.
[06:45.860 --> 06:49.460]  The passive systems get hit with a signal and they send something back.
[06:49.640 --> 06:51.060]  But there's no power source on them.
[06:52.900 --> 06:58.180]  So the project in its current form is not exactly what I had in mind when we got started.
[06:58.400 --> 07:03.820]  But we definitely dropped the in-person icebreaker element.
[07:03.820 --> 07:12.860]  But the new system allows you to put in your email and phone number and then put in your interests.
[07:13.020 --> 07:20.960]  And when your interests match up with someone else, then it will email, it will tell you that you can connect to these people.
[07:20.960 --> 07:26.940]  And then you can tell the system to send them your email address and phone number at the conference so you can meet up with them.
[07:27.900 --> 07:31.560]  So that's part of the whole website that's been built and all that stuff.
[07:33.420 --> 07:35.580]  So that's basically it.
[07:37.280 --> 07:42.260]  It took four months, which is not a lot of development time for something on this scale.
[07:42.320 --> 07:44.820]  And we're actually still debugging a lot of it today.
[07:46.180 --> 07:52.400]  We're hoping to have it running like we kind of dreamed it would work by this evening.
[07:53.240 --> 08:01.040]  So stay tuned and we're going to get the public terminals downstairs up and running so people can log in and check out if you don't have your own laptop with you.
[08:01.220 --> 08:08.860]  And so joining us up here now is Aesthetics, who's been doing the database stuff and comparing interests and things.
[08:09.600 --> 08:15.000]  And Christina, who did the website layout and has been working on visualizations.
[08:15.300 --> 08:15.620]  So...
[08:15.620 --> 08:17.900]  Can we have a DVI connector?
[08:18.160 --> 08:19.840]  No, I don't think there's DVI up here.
[08:20.300 --> 08:22.700]  If we have it in the house, somebody send a connector up.
[08:24.220 --> 08:31.620]  Aesthetics, why don't you get started and tell us a little bit about the database and the interest and all that stuff, how it's connected.
[08:32.400 --> 08:33.000]  Okay, sure.
[08:33.680 --> 08:36.940]  First off, how many people have tried to log into the website right now?
[08:37.840 --> 08:38.540]  Okay, sweet.
[08:39.360 --> 08:41.200]  And yeah, you know, we're working on some stuff.
[08:41.520 --> 08:42.940]  This is the first deployment.
[08:43.220 --> 08:45.300]  We've... I've done as much testing as possible.
[08:45.640 --> 08:47.740]  But it is the first deployment of this system.
[08:47.900 --> 08:50.220]  And as such, we're going to run into a couple little glitches.
[08:50.320 --> 08:53.520]  So it's nothing big and we're working on it.
[08:53.600 --> 08:53.820]  Yeah.
[08:54.080 --> 08:55.540]  Yeah, it's actually a really fun time.
[08:55.640 --> 08:57.760]  It's a lot less stressful than it's been for the last week or so.
[09:00.200 --> 09:08.060]  As far as the web, kind of the vision that I had when creating the database structure is looking at sites.
[09:08.400 --> 09:10.940]  What are some of the popular social networking sites?
[09:11.140 --> 09:13.380]  You've got... what are some of them?
[09:14.040 --> 09:14.440]  MySpace.
[09:14.560 --> 09:15.180]  Yeah, MySpace.
[09:15.360 --> 09:15.760]  Facebook.
[09:16.180 --> 09:16.660]  Okay, Cupid.
[09:17.380 --> 09:18.660]  You know, retarded.
[09:19.520 --> 09:20.140]  I don't know.
[09:20.460 --> 09:22.540]  How many people here use Facebook?
[09:24.620 --> 09:25.380]  Oh, come on.
[09:25.500 --> 09:25.700]  Come on.
[09:25.760 --> 09:26.480]  Don't be shy now.
[09:26.540 --> 09:28.060]  I know a lot more stories.
[09:28.520 --> 09:28.540]  Facebook.
[09:28.860 --> 09:29.100]  Get out of here.
[09:29.120 --> 09:30.040]  Confession is good for the soul.
[09:30.180 --> 09:30.540]  Come on.
[09:31.420 --> 09:33.120]  And there's also sites like LinkedIn.
[09:33.320 --> 09:35.380]  And what those do is they take people and link them together.
[09:36.620 --> 09:39.620]  Looking at different websites, like Flickr is a popular one.
[09:39.700 --> 09:42.740]  You can store images and you can assign interest tags to these images.
[09:43.260 --> 09:48.020]  And what we realized is that these images are basically objects that have interest tags assigned to them.
[09:48.540 --> 09:54.560]  And the route we took is that people can effectively be viewed as objects in this database.
[09:54.820 --> 10:00.660]  So if we have a person who logs in, and they choose interest, that's assigning interest tags to the people.
[10:00.880 --> 10:06.940]  And then with people, we can do all the really cool data mining infrastructure stuff that Flickr and all these other sites are doing.
[10:07.580 --> 10:10.480]  And additionally, we have every talk.
[10:10.980 --> 10:16.980]  Actually, one of the guys working on the data mining, Captain Fourier, he's asleep right now because he was up until like 9 a.m.
[10:17.020 --> 10:17.660]  working on this.
[10:18.120 --> 10:25.180]  He painstakingly took like six hours and went through every talk at the convention and assigned two or three specific interests to them.
[10:25.180 --> 10:29.760]  So when you log in to the website... and it'll be up soon.
[10:30.520 --> 10:39.520]  When you log in to the website, you will see what's called recommended talks, which actually takes the interest that you've selected and shows you talks that you might be interested based on that.
[10:39.580 --> 10:42.120]  And this is powerful because this convention is huge.
[10:42.240 --> 10:44.380]  There's like 103 tracks or something like that.
[10:44.780 --> 10:57.140]  And one of the biggest problems I've had with conventions is I find out after the fact that there's a talk that I wanted to see or, you know, something cool that I was interested in and I never knew about it.
[10:57.280 --> 10:59.800]  So this is actually a really cool way that we can get around that.
[10:59.980 --> 11:04.060]  And we can take the 103 talks and filter it down to like maybe 10.
[11:04.360 --> 11:07.840]  And based on that, you can actually select the talks that you're interested in.
[11:08.600 --> 11:09.100]  Let's see.
[11:10.160 --> 11:11.540]  What else do we want to know about database?
[11:11.900 --> 11:13.400]  Like there's a ton of stuff.
[11:14.920 --> 11:17.420]  What's under it kind of system-wise?
[11:18.700 --> 11:21.400]  It's PHP running on an Oracle server.
[11:22.800 --> 11:24.020]  We chose Oracle.
[11:24.340 --> 11:26.040]  We actually were using another system.
[11:26.160 --> 11:29.260]  We chose Oracle because the... Yeah, we were using MySQL.
[11:29.460 --> 11:33.620]  However, Oracle has much more powerful aggregation and data mining features.
[11:33.660 --> 11:35.580]  So we'll be able to do cool things.
[11:36.160 --> 11:44.580]  Like for example, in every track we can show the distribution of interests that are going across and we can cross-reference that with things like age or even where you're from.
[11:46.280 --> 11:52.260]  As far as privacy, I know this is going to be a really big concern because why the hell would you give us your phone number?
[11:52.420 --> 11:53.740]  I mean, seriously, right?
[11:54.780 --> 12:00.120]  The phone numbers, if you choose to enter them, are encrypted with Blowfish government strength encryption.
[12:00.220 --> 12:05.840]  So in the event that the database does get compromised, I don't see how that would happen.
[12:06.580 --> 12:08.720]  But nobody... What?
[12:10.780 --> 12:11.540]  The key is stored?
[12:11.700 --> 12:12.240]  I'm not telling.
[12:19.430 --> 12:20.290]  Is that charty?
[12:26.870 --> 12:34.910]  But basically the database and all the other stuff is on the back end on another VLAN, so there's no possible way you're getting into it.
[12:35.510 --> 12:36.770]  So have fun.
[12:37.010 --> 12:37.730]  Enjoy that.
[12:37.970 --> 12:38.330]  Yeah.
[12:38.490 --> 12:38.670]  Really.
[12:38.890 --> 12:39.650]  Seriously, guys.
[12:41.690 --> 12:42.390]  Oh, totally.
[12:42.750 --> 12:43.710]  Like... No.
[12:44.470 --> 12:45.090]  I shouldn't have here.
[12:45.290 --> 12:46.370]  Please try to hack this site.
[12:46.450 --> 12:47.190]  Let's see what happens.
[12:47.290 --> 12:49.610]  We've run AppScoun and a couple other things on it a couple times.
[12:49.610 --> 12:51.310]  And it's come out pretty well.
[12:52.170 --> 13:00.890]  If you come out... Like, in the initial run of the web vulnerabilities, we came to very little.
[13:01.190 --> 13:07.710]  So if you guys would like a challenge and want to hack the site... Do not DDOS it.
[13:08.090 --> 13:08.250]  Yeah.
[13:08.530 --> 13:09.470]  No DDoS.
[13:09.630 --> 13:09.950]  Okay?
[13:10.170 --> 13:10.710]  Get out of here.
[13:10.970 --> 13:12.410]  No DDoS attacks.
[13:12.770 --> 13:15.590]  No hardware DDoSing devices either.
[13:16.250 --> 13:17.510]  None of that is permitted.
[13:17.730 --> 13:22.070]  But if you'd like to try a creative ways to hack the site, go for it.
[13:22.130 --> 13:25.050]  And if you know Dan Kaminsky's DNS exploit, don't use that on us.
[13:29.050 --> 13:31.730]  Christina, why don't you talk about what we're trying to do with Viz?
[13:34.410 --> 13:54.950]  Well, what we were originally trying to do with Viz is kind of do a crowd mapping scenario where we track everybody by their RFID tags and using a map of each of the floors, we can visualize how many people are in each room at any given time and also what the general crowd movement is.
[13:55.790 --> 14:07.090]  Whether we have, say, congestion by the elevators, whether the rooms are packed during particular talks, and what the crowd movement down on the mezzanine is around the vendor tables.
[14:08.230 --> 14:23.970]  We wanted to add certain things like picking out a particular vendor table and seeing how many people were there within the space of an hour at any given point in time and like things like that.
[14:24.350 --> 14:28.510]  Right now we're still trying to get the visualization up because we're still implementing all the readers.
[14:29.790 --> 14:33.390]  Does anyone have a DVI to VGA connector in the house?
[14:33.390 --> 14:35.510]  Yeah, I have DVI on my Mac.
[14:35.670 --> 14:36.650]  Just bring it up.
[14:36.790 --> 14:38.030]  Yeah, that would be great.
[14:42.370 --> 14:44.250]  Hey, actually, I got one in my pocket.
[14:44.350 --> 14:44.590]  Do you?
[14:44.850 --> 14:44.990]  Yeah.
[14:46.610 --> 14:48.270]  I don't know how that happened.
[14:48.770 --> 14:50.170]  Wild party last night.
[14:52.970 --> 14:54.450]  That was totally right in my opinion.
[14:54.590 --> 14:56.690]  And you have the proper Mac one now, too.
[14:58.330 --> 15:01.030]  Damn hackers are always carrying those connectors.
[15:01.590 --> 15:01.730]  Whatever.
[15:04.010 --> 15:07.650]  Milos, why don't you talk about OpenBeacon while she's getting set up.
[15:07.670 --> 15:08.410]  Do we have projector?
[15:08.550 --> 15:09.390]  I'm not getting projector.
[15:10.570 --> 15:11.110]  We got it.
[15:11.690 --> 15:12.510]  Hey, there we go.
[15:12.630 --> 15:12.830]  Hey.
[15:14.690 --> 15:17.890]  Okay, so I will give you a brief introduction into OpenBeacon.
[15:18.730 --> 15:25.490]  The basic idea behind OpenBeacon is to create a very low cost generic HF transmitter and receiver.
[15:25.690 --> 15:30.990]  So, all the tags you currently have are also able to transmit and receive.
[15:31.230 --> 15:36.350]  So, there is a general purpose programming adapter connector for microchip CPUs.
[15:36.990 --> 15:41.330]  So, in case you are interested to update the system or create new firmware.
[15:41.470 --> 15:45.970]  For example, creating sniffer tags that can visualize the amount of people around you and stuff like that.
[15:46.510 --> 15:52.390]  Don't hesitate to come to our table and we can guide you how to change your firmware.
[15:52.390 --> 15:57.890]  So, the current implementation of the tag is pretty simple.
[15:58.130 --> 16:03.710]  It constantly transmits your ID to make sure that no one can spoof an ID.
[16:03.910 --> 16:12.130]  Every packet contains a sequence number and the whole packet is block encrypted by using XXTEA encryption.
[16:12.820 --> 16:23.330]  It's guarded by 128-bit key and it constantly transmits packets and round robins transmit strength between four different levels.
[16:24.030 --> 16:42.450]  The basic idea we did for distance estimation is to verify if, for example, a strong packet and the weak packet arrives at the given reader, then you can be pretty sure that you are near to the reader and if you get only strong packets, you are obviously far away from the reader.
[16:43.470 --> 16:52.010]  Even though we have only four power levels, we can estimate also fine-grained distance by using statistic analysis.
[16:52.790 --> 16:59.270]  So, the tags that currently are in possession transmit something around eight to ten packets a second.
[17:00.290 --> 17:06.410]  They blink something at every fourth packet or every eight packet, I'm not sure about that 100%.
[17:07.890 --> 17:13.190]  And battery will last something like three to four months constantly transmitting.
[17:13.870 --> 17:21.370]  And if you don't want to be tracked for a time because we need some time off, just remove your battery.
[17:21.650 --> 17:35.170]  So, the design criteria behind the system was that only if you need to be tracked, if you, for example, want to find out what people attending same talks, you can leave your battery in.
[17:35.290 --> 17:42.850]  But if you want to be on your own and you don't want to know where you are, you don't want to know people where you are, then you just remove battery.
[17:42.850 --> 17:52.610]  So, I think this is one of the good FID system because you are in full control of what happens with the data when data is gathered.
[17:53.290 --> 17:57.630]  And basically, you will see clearly during next days what we will do with your data.
[18:00.010 --> 18:00.530]  Yeah.
[18:00.670 --> 18:01.470]  That's basically it.
[18:01.550 --> 18:05.430]  What did you do with it at the conferences in CCC and what's different here?
[18:06.410 --> 18:06.710]  Yeah.
[18:07.270 --> 18:10.890]  What we did there was 3D visualization of the building.
[18:12.110 --> 18:21.770]  We had a guy that did an amazing 3D model of the system down to the slightest bolt in the wall.
[18:23.330 --> 18:43.750]  And basically, you were able to move virtually throughout the building and click via TAT screen on people and you could see the handle and we were able to tell where people are at a given time.
[18:43.990 --> 18:44.350]  Okay.
[18:44.970 --> 18:46.610]  Can I happen with a couple of things?
[18:46.750 --> 18:46.870]  Yeah.
[18:47.510 --> 18:49.470]  Two things I wanted to touch up upon.
[18:50.290 --> 18:55.010]  Lexicon mentioned the ping idea before and I just wanted to make sure that it's clear.
[18:55.450 --> 19:02.190]  What a ping is, is when somebody shows up like a similar user to you who has a shared interest, you see two icons next to their name.
[19:03.090 --> 19:07.670]  One of them is a picture of an email and the other is like this cloud that's supposed to be for SMS.
[19:07.910 --> 19:18.850]  And what happens when you click on those icons, if you click on the email, the system will send them an email with your email address in it and for the cell phones, it will send them a text message with your cell phone number in it.
[19:18.850 --> 19:19.890]  And that's all it does.
[19:20.250 --> 19:24.330]  And it also sets a flag in the database so you can only actually send them a ping once.
[19:24.430 --> 19:25.850]  So you can't like mail bomb somebody.
[19:26.010 --> 19:28.830]  That was a really big concern, trying to write all the mail functions.
[19:29.310 --> 19:33.670]  The other thing which I suspect... did you touch upon the database cleansing at all?
[19:34.090 --> 19:34.750]  No, no.
[19:34.750 --> 19:35.550]  I haven't talked about database.
[19:35.610 --> 19:35.870]  Okay.
[19:35.950 --> 19:36.750]  Are we gonna get to that later?
[19:37.010 --> 19:37.410]  Or...
[19:38.610 --> 19:39.370]  We can?
[19:39.870 --> 19:41.250]  I can talk about it right now.
[19:41.730 --> 19:41.910]  Yeah.
[19:42.250 --> 19:45.690]  After this project, basically all the source code is gonna be GPL'd.
[19:47.150 --> 19:52.290]  And regards to the data, because this is a really cool opportunity to do experiments on it.
[19:52.370 --> 19:56.790]  I'm sure universities and whatnot might be interested in seeing precisely what we can do when we cross...
[19:56.790 --> 19:57.530]  Taking offers?
[19:57.910 --> 19:58.250]  Yeah.
[20:00.550 --> 20:10.030]  But when we cross-reference like interests and details, like where somebody is from with the talks that they went to and just migration habits, you can form some really cool ideas out of that.
[20:10.930 --> 20:12.870]  There's also a big concern about privacy.
[20:13.190 --> 20:18.930]  So things that we consider personal information would be like a phone number, for example.
[20:19.070 --> 20:20.050]  What else would we consider that?
[20:20.170 --> 20:20.250]  Yeah.
[20:20.270 --> 20:21.990]  Phone number and the handle and...
[20:21.990 --> 20:22.230]  Yeah.
[20:22.390 --> 20:22.990]  And the handle.
[20:23.090 --> 20:26.470]  Basically, we saw what happened when AOL leaked all that data.
[20:26.610 --> 20:28.050]  Didn't they get broken into or something?
[20:28.210 --> 20:29.350]  And people were able to take...
[20:29.350 --> 20:29.910]  Which time?
[20:30.410 --> 20:30.550]  Yeah.
[20:30.690 --> 20:30.770]  Yeah.
[20:31.150 --> 20:32.870]  This isn't the 90s, but...
[20:32.870 --> 20:33.910]  This is the 90s.
[20:34.290 --> 20:40.150]  But people were able to take searching habits and infer exactly who that was based on that.
[20:40.250 --> 20:44.610]  So we're trying to strip away as much personal data as possible and still make it a valuable resource.
[20:44.950 --> 20:50.590]  So I know there's a very big privacy concern and I want to address the fact that we definitely know about that.
[20:50.590 --> 20:52.070]  But we're doing everything we can.
[20:52.230 --> 20:54.690]  Let's go back to Christina with the website stuff.
[20:57.230 --> 20:58.130]  Is there anything you want me to touch on?
[20:58.130 --> 20:58.570]  Do you have a mic?
[20:59.090 --> 20:59.230]  What?
[20:59.410 --> 20:59.610]  Yeah.
[21:00.150 --> 21:01.770]  Anything you want me to touch on in particular?
[21:01.930 --> 21:02.790]  Just do a walkthrough of it.
[21:02.930 --> 21:03.130]  No.
[21:03.230 --> 21:06.050]  Just show the system and what you think is interesting about it, I think.
[21:06.330 --> 21:06.630]  All right.
[21:07.570 --> 21:09.430]  Well, I'll show what you can see.
[21:09.770 --> 21:10.130]  Thanks.
[21:10.650 --> 21:13.510]  I'll show what you can see without being logged in.
[21:13.570 --> 21:15.050]  Obviously, here's the login screen.
[21:16.430 --> 21:26.470]  If we go to the schedule page, what we have here is the schedule set up by room for each day.
[21:27.830 --> 21:31.610]  So you can see it and say you want to view talk detail.
[21:33.090 --> 21:35.350]  I don't think all the talk details are in.
[21:37.890 --> 21:39.630]  I think it's part of the...
[21:39.630 --> 21:41.830]  There's a database thing that Michael's working on right now.
[21:42.010 --> 21:42.130]  Oh, okay.
[21:42.870 --> 21:43.150]  All right.
[21:43.230 --> 21:45.570]  So the database is being worked on, but...
[21:45.570 --> 21:51.470]  When you open that, that will show the title of the talk, the abstract of it, and then the speaker and the speaker bios for everyone who's in it.
[21:51.870 --> 21:52.090]  Yeah.
[21:52.590 --> 21:53.790]  So that's how it's supposed to work.
[21:55.250 --> 21:57.030]  We have frequently asked questions.
[21:59.750 --> 22:05.010]  Our visualize page, again, we're still configuring the readers for the second floor.
[22:05.010 --> 22:20.710]  What we were planning to have was this map that you can view, and then over it will be a little real-time visualizer with basically dots that represent each user with a tag moving around the space.
[22:21.030 --> 22:26.050]  And we have a lot of readers down there, so this will be a little simpler.
[22:26.050 --> 22:34.310]  On the 18th floor, we're mostly doing weather map style, like room concentration levels.
[22:34.490 --> 22:44.050]  So say you click on Engressia, and you get a room detail, what's going on in the room, who's doing the talk, and you'll be able to get how many people are in the room at the time.
[22:46.530 --> 22:48.730]  If you want to log in...
[22:48.730 --> 22:50.150]  Can I use a test log in for this?
[22:51.510 --> 22:53.950]  To show them how the schedule looks when they're logged in?
[22:56.690 --> 22:58.570]  I took out the test account, I think.
[22:58.770 --> 22:59.070]  All right.
[22:59.070 --> 23:00.190]  Do you just log in with my...
[23:00.190 --> 23:00.590]  Yeah.
[23:02.950 --> 23:03.830]  Let's see...
[23:06.560 --> 23:07.480]  Let's see if that goes.
[23:07.620 --> 23:07.900]  There you go.
[23:07.960 --> 23:08.200]  All right.
[23:08.400 --> 23:16.180]  So now when you're logged in, you have a bunch of contextual information that is unique to you.
[23:16.400 --> 23:20.520]  You can view your schedule, what talks you picked.
[23:23.640 --> 23:32.800]  Well, when you create a new account, you put in your personal information and you pick your interests and you get recommended talks based on what those interests are.
[23:33.320 --> 23:38.670]  And what you can also do, let me show you the schedule, is you can check off...
[23:39.130 --> 23:41.410]  Now there's check boxes now that you're logged in.
[23:41.530 --> 23:44.090]  You can check off which talks you want to see.
[23:44.350 --> 23:45.530]  Click update schedule.
[23:45.770 --> 23:49.590]  And then you'll be sent reminders to go to those talks that you're interested in.
[23:52.690 --> 23:54.790]  Also, now you can view...
[23:54.790 --> 23:58.930]  Since you're logged in, you can view new users that have just logged in.
[23:58.930 --> 24:01.950]  Or you can choose interests in your profile.
[24:07.440 --> 24:09.920]  And basically just a little bit of personal data.
[24:10.200 --> 24:11.240]  You choose your interest.
[24:11.480 --> 24:12.040]  You submit.
[24:14.460 --> 24:16.260]  Can I change your account info?
[24:16.440 --> 24:16.780]  Go for it.
[24:16.880 --> 24:17.140]  All right.
[24:17.140 --> 24:17.160]  Yeah.
[24:21.880 --> 24:24.760]  So, say I just pick five random things.
[24:29.170 --> 24:29.650]  Ah.
[24:31.470 --> 24:38.970]  See, we have email...
[24:38.970 --> 24:39.630]  Interesting.
[24:42.230 --> 24:42.710]  And...
[24:43.510 --> 24:45.690]  Yeah, as you can see, this is still...
[24:45.690 --> 24:47.570]  We're still patching up some bugs in it.
[24:48.170 --> 24:48.590]  Yeah.
[24:48.590 --> 24:48.610]  Yeah.
[24:48.770 --> 24:49.090]  That's cool.
[24:49.230 --> 24:51.110]  That's one I didn't know about, so...
[24:52.110 --> 24:55.590]  Well, now we know there's accurate checking.
[24:56.450 --> 24:58.670]  What's the SMS system you're using, Kristina?
[25:01.430 --> 25:03.270]  Aesthetics can answer that better than I can.
[25:03.470 --> 25:08.090]  The SMS system that we're using, the provider drop-down is inherited from...
[25:08.090 --> 25:11.490]  There's a table in the database with the provider name and email addresses.
[25:11.490 --> 25:21.270]  And if you put in your cell phone number at an email address.com, for example, you know, 1234 at, what, text.att.net, something like that.
[25:21.410 --> 25:26.310]  And you send an email, the system will actually take that and forward it as a text message to a cell phone.
[25:26.410 --> 25:27.630]  So, that's how that's working.
[25:27.910 --> 25:29.370]  And it's actually really simple to do.
[25:29.910 --> 25:32.770]  Weren't we using a new system like TextML or something?
[25:32.970 --> 25:33.970]  Oh, that...
[25:33.970 --> 25:34.930]  Oh, it was abandoned?
[25:35.150 --> 25:35.370]  Yeah, squashed.
[25:35.730 --> 25:35.870]  Okay.
[25:36.830 --> 25:37.430]  That I knew.
[25:37.510 --> 25:37.690]  Okay.
[25:38.590 --> 25:40.910]  It didn't really work for what we were trying to implement.
[25:40.910 --> 25:44.710]  This was very simple and it's much easier to error troubleshoot, so...
[25:44.710 --> 25:45.050]  Okay.
[25:46.590 --> 25:47.650]  Do you have anything else?
[25:50.550 --> 25:55.390]  We're going to have games, but those are dependent on the RFID readers gathering data.
[25:55.630 --> 25:58.150]  Again, that's going to be dependent on that setup being finished.
[25:59.950 --> 26:02.530]  So, it's not quite ready yet.
[26:03.270 --> 26:11.750]  But the game is going to be based on where you are, how many talks you attend, you have a score that counts down.
[26:12.430 --> 26:14.950]  And the more talks you attend, the more points you get, basically.
[26:17.630 --> 26:21.910]  And at the end of the weekend, whoever has the most points, like, wins some prizes.
[26:24.030 --> 26:24.430]  Okay.
[26:24.650 --> 26:30.610]  So, Matt was in charge of the infrastructure with Rage Fury, who's at work right now.
[26:30.910 --> 26:34.570]  So, Matt, why don't you say a little bit about what we had to build?
[26:34.570 --> 26:34.690]  Yep.
[26:35.070 --> 26:35.290]  Okay.
[26:35.630 --> 26:36.890]  So, infrastructure.
[26:37.510 --> 26:41.210]  My co-lead is actually at work because he has a job.
[26:41.350 --> 26:41.890]  What a concept.
[26:43.450 --> 26:53.390]  Anyway, since I am not at work, I am here to talk to you about the infrastructure and what is different here that was not implemented over at the CCC in Germany.
[26:54.470 --> 27:10.130]  The biggest portion of this that's the difference is the triangulation and, of course, the social networking aspect, which was mostly controlled by the databasing and our data mining guy, Captain Fourier.
[27:10.370 --> 27:11.390]  I need to mention this guy.
[27:12.450 --> 27:14.110]  Incredible data miner.
[27:14.110 --> 27:16.970]  And he's definitely going to be a PhD.
[27:17.210 --> 27:17.650]  Absolutely.
[27:18.250 --> 27:24.350]  So, in any case, a few challenges.
[27:24.770 --> 27:27.890]  Getting enough people to wire things up.
[27:28.130 --> 27:36.810]  Getting enough people to crimp cables and put readers into the particular places so that there can be some proper triangulation.
[27:37.010 --> 27:44.490]  We didn't go with another system like quadrangulation because we don't need to know where people are within two inches of where they actually are.
[27:44.880 --> 28:01.090]  So, we just used some basic triangulation to give a general overview of where the person is according to area, as you've seen on the screen when Christina showed you the areas like Turner and Hopper and all the different conference areas.
[28:03.670 --> 28:09.650]  In any case, the system setup was rather simple.
[28:10.550 --> 28:14.070]  Basically, you need a couple of elements.
[28:14.450 --> 28:21.170]  One being that you have Linux, of course, or any type of UNIX operating system.
[28:22.370 --> 28:24.950]  And you have an aggregation system.
[28:25.770 --> 28:45.090]  The aggregator will then send all the data from the readers and it will just commit it to the database after the aggregation process takes place and the triangulation is then turned into coordinates, which then we can get reads on where people are.
[28:45.270 --> 28:49.450]  Why can't we just take the data that's straight from the readers and put it into the database?
[28:49.650 --> 28:51.010]  Why does it have to go through something?
[28:51.750 --> 28:55.770]  Because the data that comes out of the readers is not English.
[28:55.770 --> 29:04.190]  So we needed to switch it to actual coordinates so we can get inches and feet and where people are.
[29:04.570 --> 29:07.430]  So it's combining the information from multiple readers?
[29:07.670 --> 29:08.510]  Yes, that's correct.
[29:08.950 --> 29:09.370]  That's correct.
[29:11.190 --> 29:14.070]  Does anyone have any questions about the infrastructure?
[29:14.310 --> 29:18.510]  I think there's a microphone over there if anyone wants to line up and hit us.
[29:19.930 --> 29:21.730]  There's a lot of implications about the...
[29:21.730 --> 29:23.830]  Please go to the microphone if you have a question.
[29:26.130 --> 29:27.330]  That is a microphone, right?
[29:36.280 --> 29:37.040]  Is it on now?
[29:43.070 --> 29:43.630]  All right.
[29:44.310 --> 29:49.050]  As far as the hardware setup, how much did each one of the receivers cost?
[29:49.330 --> 29:50.310]  And about how many are there?
[29:51.490 --> 29:52.490]  Millers can answer that.
[29:52.870 --> 29:54.890]  How much did the what cost?
[29:55.090 --> 29:55.730]  The receivers?
[29:56.070 --> 29:56.350]  The readers.
[29:56.350 --> 29:56.990]  Oh, the readers.
[29:57.210 --> 29:57.430]  The readers.
[29:57.590 --> 29:58.230]  Yeah, Milos.
[29:58.510 --> 30:00.590]  The readers are custom made by us.
[30:00.730 --> 30:03.270]  They are not publicly sold yet.
[30:03.690 --> 30:09.730]  But the estimated cost will be something like $200 to $250, something like that.
[30:10.050 --> 30:12.950]  They basically provide power Ethernet support.
[30:12.950 --> 30:19.610]  So they are pretty easy to distribute around the facility because you only have to connect the Ethernet cable.
[30:19.990 --> 30:22.690]  And they basically are very stupid.
[30:22.690 --> 30:30.630]  They just pick up stuff that comes in on RF side and blindly forwards the encrypted packets to the aggregator.
[30:31.270 --> 30:34.170]  And only the aggregator knows the secret key.
[30:34.350 --> 30:41.370]  And there they are decrypted and forwarded to all clients that will do, for example, the position estimation and so on.
[30:41.370 --> 30:48.710]  And just for information, for community purposes, if you want to play around with this technology, there is a much cheaper possibility.
[30:49.050 --> 30:53.970]  We have also a very simple design which is based on USB.
[30:54.550 --> 30:56.110]  It's an open beacon USB project.
[30:56.770 --> 30:59.790]  And this is something like $85 or so.
[31:00.050 --> 31:02.730]  And you can freely change the firmware.
[31:02.930 --> 31:05.170]  It's a 32-bit ARM processor.
[31:05.530 --> 31:08.890]  And you have a serial interface over USB.
[31:09.150 --> 31:11.970]  So you can have your reader basically...
[31:12.650 --> 31:15.930]  You can turn your laptop to a reader and you can walk around the building.
[31:15.930 --> 31:23.690]  So in case you're interested to play around with this technology, you can lend some of the readers at the AMD project from us.
[31:24.010 --> 31:28.170]  And you can, for example, try to write your own software for your laptop.
[31:28.550 --> 31:32.530]  And if you have any interesting ideas, please share them with us.
[31:32.970 --> 31:35.610]  We have about 50 readers throughout the building, I think.
[31:35.670 --> 31:36.010]  Is that right?
[31:36.410 --> 31:36.630]  Yeah.
[31:36.850 --> 31:38.450]  Currently, there are around 50 readers.
[31:39.050 --> 31:41.330]  And where can they get the USB?
[31:41.590 --> 31:43.710]  Is it on a website to order them?
[31:43.710 --> 31:44.730]  Yeah.
[31:45.010 --> 31:47.170]  On the openbeacon.org website there.
[31:47.390 --> 31:49.870]  You can also download the schematics.
[31:50.890 --> 31:54.670]  And if you want, you can build your own design based on it.
[31:55.010 --> 32:00.630]  The design is published under a Creative Commons license.
[32:01.570 --> 32:03.610]  And basically, you get everything there.
[32:03.770 --> 32:06.610]  The firmware is licensed under a GPL license.
[32:07.290 --> 32:10.170]  You don't have to pay for any kind of programming equipment.
[32:10.630 --> 32:11.790]  It's unbrickable.
[32:11.790 --> 32:13.430]  You can program it over USB.
[32:13.730 --> 32:18.090]  So I think the initial costs are basically only the device.
[32:18.870 --> 32:19.430]  So...
[32:19.430 --> 32:20.030]  Yeah.
[32:20.310 --> 32:20.670]  Next.
[32:21.510 --> 32:22.250]  Hi, guys.
[32:22.330 --> 32:23.690]  I have a two-part question.
[32:23.950 --> 32:26.530]  The first one is, did you have any problem?
[32:26.570 --> 32:33.250]  And how did you find a solution to discriminating between floors above and below you in this case?
[32:33.250 --> 32:40.210]  And the second part is, did you have trouble filtering out any ambient noise that was maybe outside of your project area?
[32:41.370 --> 32:42.390]  Yeah, it's...
[32:42.390 --> 32:44.790]  I might be able to address part of the first...
[32:44.790 --> 32:45.750]  Yeah.
[32:46.390 --> 32:46.890]  Sorry.
[32:47.030 --> 32:49.250]  I might be able to address the part of the first...
[32:49.250 --> 32:50.870]  I was working on the visualization part.
[32:51.990 --> 32:56.030]  Originally, we had a design that would have readers all over the building.
[32:56.030 --> 32:59.290]  So we were expecting to have a sort of cube or a volume.
[32:59.290 --> 33:02.090]  And in that case, we would have used quadrangulation.
[33:02.730 --> 33:09.270]  However, as time progressed and as time kind of diminished, we decided to go with the triangulation system.
[33:09.530 --> 33:17.370]  So it is possible for someone, for example, to be on the 19th floor or on the roof or the floor below and penetrate through the floor itself.
[33:18.870 --> 33:19.610]  It'll happen.
[33:19.750 --> 33:20.230]  It'll appear.
[33:21.430 --> 33:23.490]  But we just assumed a flat model.
[33:24.450 --> 33:27.250]  Yeah, but there is a very simple solution for this.
[33:27.730 --> 33:32.510]  So as long as you have at least one reader on each floor, you can discriminate it.
[33:32.570 --> 33:46.050]  So if you get a lot of weak packets on one floor, then you know that if you pick up only strong packets on the floor below or above, then you are sure that you are not on the floor below or above.
[33:46.050 --> 33:53.670]  So just, you have to have to make sure that in the underlying floors, you have additional reader at least.
[33:53.890 --> 33:58.030]  And then you can easily discriminate other rooms and floors.
[34:02.680 --> 34:04.560]  Yeah, and addressing the noise.
[34:05.260 --> 34:07.940]  The system actually works at 2.4 gigahertz.
[34:08.140 --> 34:14.380]  So what we basically did, we moved our frequency to the upper end of the ISM band.
[34:14.380 --> 34:18.160]  It's currently at 2.4.8.1 megahertz.
[34:18.860 --> 34:22.160]  And there you have very few noise.
[34:22.320 --> 34:25.020]  It's inside the guard band of Bluetooth.
[34:25.460 --> 34:31.580]  And it's basically you get only some side bands of wireless LAN.
[34:31.720 --> 34:33.180]  So we are pretty safe with this.
[34:33.180 --> 34:40.800]  And even if single packets break, the system is constantly transmitting something like eight packets a second.
[34:40.900 --> 34:42.760]  So if you lose some, it's not a big deal.
[34:43.720 --> 34:45.480]  We can discriminate it over time.
[34:46.320 --> 34:48.160]  So we don't regard this as a problem.
[34:48.200 --> 34:52.000]  And we currently didn't see any problems with this.
[34:52.220 --> 34:58.060]  For example, CCC, there are something like 2,000 people in a much tighter space like here with laptops.
[34:58.060 --> 35:00.840]  And also a transmit 2.4 gigahertz.
[35:00.980 --> 35:02.320]  And we never had problems with this.
[35:02.700 --> 35:04.440]  It's 2 megabits, by the way.
[35:04.520 --> 35:06.560]  So the on-air time is very, very short.
[35:06.760 --> 35:09.020]  So it's hard to jam it, actually.
[35:10.720 --> 35:11.900]  So you're welcome to try.
[35:14.980 --> 35:26.140]  I've had a lot of fun reading through the implementation of the position estimation for previous iterations of projects like this with OpenBeacon.
[35:26.140 --> 35:31.640]  And aesthetics, I believe we talked on IRC a few weeks ago.
[35:31.760 --> 35:37.220]  And you said you had some fancy new way of doing the position estimation.
[35:38.600 --> 35:44.080]  So can you guys provide any information about that new system?
[35:44.800 --> 35:46.200]  That's a rage-free question.
[35:46.600 --> 35:48.780]  Yeah, he's not here.
[35:49.040 --> 35:50.180]  Do you have information about that?
[35:50.360 --> 35:52.660]  Yeah, that is definitely a rage-free question.
[35:52.660 --> 36:01.440]  Rage-free, my co-lead, actually did a lot of the visualization work as well for his position estimation.
[36:02.300 --> 36:05.240]  So he's the real genius behind that one.
[36:05.620 --> 36:10.620]  So when rage-free gets into the con, we can forward him on to you.
[36:10.900 --> 36:11.920]  Do you have a name?
[36:13.540 --> 36:14.080]  Jeremy.
[36:14.800 --> 36:15.120]  Jeremy?
[36:15.340 --> 36:15.440]  Okay.
[36:15.440 --> 36:22.020]  If you don't happen to find rage-free during this weekend, you can email projects at hope.net.
[36:22.180 --> 36:24.940]  That comes to me and I will forward it to him, the questions.
[36:25.840 --> 36:26.320]  Okay.
[36:27.420 --> 36:28.500]  Yeah, that's true.
[36:31.700 --> 36:44.920]  I'm worried that this might be a related question and you may not be able to answer it, but I'm curious if you could say some more about the coordinate system and how you lay things out, how you measured where things are.
[36:44.920 --> 36:45.920]  Same problem?
[36:46.860 --> 36:47.260]  No.
[36:47.760 --> 36:47.940]  Okay.
[36:48.220 --> 36:48.420]  Cool.
[36:48.760 --> 36:48.840]  Cool.
[36:50.760 --> 36:50.980]  Yeah.
[36:51.700 --> 36:52.980]  There are basically two approaches.
[36:53.240 --> 36:58.380]  So in rooms like this, basically you see four readers in corners.
[36:58.540 --> 37:05.960]  So we try to manage to see people with as few as possible meat in between.
[37:05.960 --> 37:08.440]  So therefore they are usually pretty high.
[37:09.180 --> 37:11.440]  And here we can do actual triangulation.
[37:11.800 --> 37:18.260]  But in other rooms which are not so important, we basically have, for example, in the far out room, I don't know the name by heart.
[37:18.260 --> 37:18.720]  Madison.
[37:19.780 --> 37:20.660]  Or Susie.
[37:20.900 --> 37:21.140]  Yeah.
[37:21.380 --> 37:22.080]  The Susie room.
[37:22.520 --> 37:23.980]  That we have only a single reader.
[37:24.500 --> 37:26.200]  And the idea is very simple.
[37:26.600 --> 37:34.820]  If you're in one room, get strong and weak packets, and in the other rooms you get only strong packets, then you are pretty sure that a person is in a single room.
[37:34.820 --> 37:38.420]  So basically you see two concepts widely combined.
[37:38.940 --> 37:40.940]  Some areas are only covered as areas.
[37:41.140 --> 37:48.300]  So we only know that you are in the particular area and you are not in the area next to it.
[37:48.720 --> 37:58.080]  But for, I think, most of the conference rooms and in the main hall, in the mezzanine level, we can do triangulation.
[37:58.260 --> 38:04.080]  Because you can be always seen by at least three to four readers.
[38:04.080 --> 38:13.720]  And what is very important for triangulation system that you don't have any obstacles in the line of sight between antennas.
[38:13.900 --> 38:20.320]  So where you have all these pillars in the mezzanine level below, they make triangulation very, very difficult.
[38:20.640 --> 38:29.420]  So therefore we limited triangulation to basically all open spaces and do area-based estimation for all other spaces.
[38:29.420 --> 38:30.960]  So it's a mixed approach, in fact.
[38:31.940 --> 38:35.800]  Yeah, this building is actually built directly over the Amtrak lines.
[38:36.040 --> 38:37.720]  And it's quite a marvel of engineering.
[38:38.040 --> 38:40.480]  But there is a hell of a lot of steel in the structure.
[38:40.700 --> 38:46.240]  So going through walls with any sort of signal inside the Hotel Pennsylvania is quite difficult.
[38:49.780 --> 38:52.340]  What's, like, the maximum capacity for one of these beacons?
[38:52.520 --> 38:55.440]  Meaning, like, how many RFIDs can it read at a single point in time?
[38:57.880 --> 39:06.380]  The last time we tried, a single Ethernet reader can pick up something like six to eight hundred packets a second easily.
[39:06.600 --> 39:07.660]  We haven't tried more.
[39:08.220 --> 39:09.820]  And same for open beacon USB.
[39:09.820 --> 39:13.620]  So I think it can't be maxed out currently.
[39:14.260 --> 39:20.500]  The only, say, limitation is that the readers only have a ten megabit Ether interface.
[39:20.800 --> 39:25.040]  But I don't think that it hurts because it immediately goes into a hundred megabit switch.
[39:25.040 --> 39:27.360]  And from there to a gigabit switch.
[39:29.400 --> 39:33.960]  So, for each single line, the ten megabit doesn't count.
[39:34.140 --> 39:41.240]  Because on air you can have, anyway, only two megabits aggregated bandwidth.
[39:42.100 --> 39:45.300]  There is one thing that may be an interesting figure to know.
[39:45.820 --> 39:51.540]  We tried how easy it is to break it by transmitting a lot of packets at the same time.
[39:51.700 --> 39:53.940]  And we tried to measure actual collisions.
[39:53.940 --> 40:00.500]  So, we transmitted something like one thousand six hundred packets at a time.
[40:00.760 --> 40:02.660]  And there were actually very few collisions.
[40:02.940 --> 40:05.620]  Something like few percents or so.
[40:06.280 --> 40:07.260]  So, yeah.
[40:07.520 --> 40:25.600]  The basic algorithm behind avoiding collisions is we wait for a random time based on a software, a random number generator that is fed by a secret random seed and your tag ID.
[40:26.940 --> 40:32.580]  So, basically, every tag waits for a random time until it retransmits its ID.
[40:32.760 --> 40:35.600]  And that's basically our approach to avoid collisions.
[40:35.780 --> 40:41.840]  So, even if two tags collide, then the probability that on the next transmission they will collide again is very low.
[40:41.840 --> 40:47.680]  And given the very short on-air time, which is basically now 16 bytes at two megabits.
[40:47.880 --> 40:50.920]  So, we are looking at something like one and a half milliseconds or so.
[40:51.140 --> 40:56.000]  On air time, the probability is very low for a collision.
[40:56.220 --> 41:00.860]  And if there is a collision, it doesn't hurt because the next packet will come very soon.
[41:02.040 --> 41:02.800]  Very nice.
[41:02.940 --> 41:03.040]  Thanks.
[41:07.130 --> 41:07.690]  Yeah.
[41:07.770 --> 41:14.190]  I was just wondering if you had two tags or multiple tags that transmitted the same ID, what do you do with that?
[41:15.410 --> 41:20.310]  If we will see this, then I would definitely wonder how this could happen.
[41:20.510 --> 41:26.810]  Because the one person that wants to do so has only two options.
[41:26.950 --> 41:31.410]  One option is to sniff data from another tag and retransmit it.
[41:31.410 --> 41:39.350]  But, then we see it on the incrementing sequence number for each packet that someone duplicated packets.
[41:39.570 --> 41:41.430]  So, we can filter them out, for example.
[41:41.550 --> 41:47.690]  So, for example, if we see a given sequence number, we just don't allow sequence number below the sequence number.
[41:47.830 --> 41:49.750]  So, you can't do a replay attack.
[41:49.950 --> 41:52.650]  So, that's, in my opinion, it's not possible.
[41:53.070 --> 41:57.970]  And the other approach is you have to know the 128-bit X60A key.
[41:58.970 --> 42:02.790]  So, I think you don't have enough time to brute force it during the conference.
[42:03.250 --> 42:06.310]  What if you intentionally programmed two tags with the same ID?
[42:06.710 --> 42:07.750]  That's definitely possible.
[42:08.010 --> 42:10.230]  For example, the USB tag is freely programmable.
[42:10.610 --> 42:11.190]  This one, too.
[42:11.370 --> 42:13.210]  So, you only need to know the key.
[42:13.710 --> 42:15.690]  And then you can program a new firmware.
[42:16.750 --> 42:18.010]  We haven't tried it, actually.
[42:18.230 --> 42:19.950]  So, maybe the manufacturer did a problem.
[42:20.130 --> 42:21.890]  Maybe the copy protection bit isn't set.
[42:22.790 --> 42:24.730]  But we have to verify this.
[42:24.850 --> 42:25.790]  We haven't verified this yet.
[42:25.950 --> 42:28.630]  So, the copy protection bit is supposed to be set.
[42:28.810 --> 42:30.790]  So, you are supposedly not...
[42:31.610 --> 42:39.690]  It's not possible to read out the microchip processor and get hold of the key.
[42:39.950 --> 42:41.790]  So, I think without a key, you are pretty lost.
[42:44.110 --> 42:45.030]  Any more questions?
[42:45.030 --> 42:46.650]  The key is really random.
[42:47.410 --> 42:50.730]  It's not a bad read from a text or something like that.
[42:50.950 --> 42:52.490]  It's from dev view random.
[42:52.690 --> 42:54.170]  So, keep trying.
[42:55.890 --> 42:56.410]  Yeah.
[42:59.250 --> 42:59.930]  Go ahead.
[43:00.470 --> 43:00.970]  Just curious.
[43:01.170 --> 43:02.470]  Who manufactured the tag?
[43:05.530 --> 43:06.570]  Chinese manufacturer.
[43:06.890 --> 43:07.050]  Yeah.
[43:07.190 --> 43:22.090]  This time, Mitch from Cornfield Electronics, the guy you know for sure, the one that created TV-B-Gone pointed out his Chinese contact which also built the TV-B-Gone devices.
[43:22.790 --> 43:27.510]  And we had production over there basically to decrease costs.
[43:27.890 --> 43:29.310]  So, it went well.
[43:29.490 --> 43:32.270]  So, we are pretty lucky about this hint.
[43:32.830 --> 43:39.410]  And the tags before for the CCC Congress were basically built at a Berlin based company.
[43:40.390 --> 43:45.530]  And the ones that are not active, are they passive?
[43:45.830 --> 43:46.470]  The other tags?
[43:46.750 --> 43:47.250]  No, no.
[43:47.370 --> 43:50.830]  They are just printed circuit boards without the circuits attached.
[43:52.210 --> 44:03.190]  So, in case you are interested to build these boards by yourself, also from the previous tags and I think we will also release the Gerber files for these tags too.
[44:03.190 --> 44:10.890]  So, you can download Gerber files for the previous tags and these tags from the OpenBeacon website and can have them built by yourself.
[44:11.090 --> 44:12.200]  It is also open design.
[44:13.110 --> 44:16.150]  So, you are free to do whatever you want with them.
[44:16.310 --> 44:21.670]  Just download Gerber files, build them in any quantity and do whatever is fun.
[44:21.870 --> 44:23.070]  But tell us about it.
[44:23.250 --> 44:26.010]  Because most people just use it and don't tell us about it.
[44:26.010 --> 44:28.530]  And so, we are very curious what you actually do with this stuff.
[44:29.510 --> 44:30.790]  So, happy hacking.
[44:31.830 --> 44:32.350]  Skirker?
[44:33.030 --> 44:38.010]  What would happen if you jammed the frequency with the high powered transmitter?
[44:39.750 --> 44:40.270]  Yes.
[44:40.810 --> 44:46.890]  So, you would definitely locally break it but you have to apply the power constantly.
[44:46.890 --> 45:01.630]  So, even if with wireless LAN it's difficult because you have in normal protocols you have this approach where the wireless LAN interface listens for traffic and then starts transmitting.
[45:02.050 --> 45:06.150]  So, wireless LAN also tries to avoid collisions.
[45:06.510 --> 45:08.510]  So, yes, you could jam it I think.
[45:08.810 --> 45:09.710]  But, oh, yes.
[45:09.890 --> 45:11.370]  It's hard to match the frequency.
[45:11.570 --> 45:14.770]  So, you will have to patch your driver I think at least.
[45:14.770 --> 45:20.510]  So, to move your center frequency up to 2, 4, 8, 3 megahertz.
[45:20.930 --> 45:22.210]  We've got the crystals now.
[45:23.370 --> 45:23.890]  Pardon?
[45:24.430 --> 45:25.930]  We've got the crystals to do it.
[45:26.870 --> 45:28.890]  No, the crystals won't help at all.
[45:29.030 --> 45:33.250]  These are 16 megahertz crystals and there is a cool PLL inside.
[45:33.570 --> 45:42.550]  So, but I think with any recent EtherOS driver you can, I guess, move your center frequency up to 2, 4, 8, 3 megahertz.
[45:43.650 --> 45:44.510]  Okay, last question.
[45:44.850 --> 45:49.510]  One last thing I want to add on to the question which is, if you do it, we'll notice.
[45:50.510 --> 45:52.650]  Our visualization is designed to follow people.
[45:52.770 --> 45:56.170]  The minute we see a big gaping hole somewhere, we're going to know somebody's playing a little trick.
[45:57.050 --> 45:57.990]  You're invited to do it.
[45:58.190 --> 45:59.070]  I'd like to see it.
[46:00.330 --> 46:01.170]  But, good luck.
[46:01.350 --> 46:04.090]  Yeah, we're curious to see what that's going to look like if somebody tries it.
[46:05.510 --> 46:06.470]  Estimated cost?
[46:07.270 --> 46:07.670]  The whole thing?
[46:08.470 --> 46:09.350]  Estimated cost?
[46:09.410 --> 46:09.910]  For the whole thing.
[46:10.630 --> 46:14.310]  There was so much donated in hardware terms.
[46:14.550 --> 46:16.470]  We were borrowing so many boxes.
[46:16.610 --> 46:18.310]  Rage Fury supplied about half the equipment.
[46:18.650 --> 46:20.990]  Milos is loaning us all the readers.
[46:20.990 --> 46:29.890]  And we got the cable donated by a guy named Michael from Long Island who has an RFID company out there.
[46:30.970 --> 46:36.170]  And the badges themselves, Emmanuel pitched in about 12 grand to get them bought.
[46:36.330 --> 46:40.310]  So these things are about $7, a little over $7 a piece without the battery in them.
[46:41.070 --> 46:44.750]  Not only that, but the system back ends that we've used.
[46:46.130 --> 46:48.430]  Did you want to include time with that cost?
[46:49.610 --> 46:49.650]  And then hours.
[46:49.650 --> 46:53.670]  Or in case you wanted to, like, corporatizing the entire system.
[46:53.930 --> 46:54.530]  After hours work.
[46:54.670 --> 46:55.430]  After hours work.
[46:55.430 --> 46:55.730]  After hours work.
[46:55.730 --> 46:57.090]  Programming hours from these guys.
[46:57.410 --> 46:57.850]  It's like...
[46:57.850 --> 46:59.810]  18 hours a day kind of thing.
[47:00.950 --> 47:06.430]  I was actually looking into if you want to actually do this on an enterprise level.
[47:06.430 --> 47:12.450]  However, there are many applications that this can be used for on an enterprise level.
[47:13.090 --> 47:20.870]  So, if you were looking at it from that perspective, you'd probably want to look at how large the area is.
[47:21.170 --> 47:24.530]  And how many people are going to be in that particular area.
[47:25.150 --> 47:32.430]  For the project, we used an Oracle backend because of the number of hits we'd be getting at all times.
[47:32.430 --> 47:37.310]  And it has the best aggregation and data mining capability, in my opinion.
[47:38.250 --> 47:43.530]  So, frankly, the cost really is a variable.
[47:43.530 --> 47:44.730]  And it changes.
[47:46.050 --> 47:51.590]  So, for this particular zone, $15,000.
[47:52.130 --> 47:53.770]  I have one last question.
[47:53.970 --> 47:54.930]  And that's for Milos.
[47:55.270 --> 47:58.470]  What is the purpose of the button that you put in the system?
[47:58.470 --> 47:59.170]  Ah, yeah.
[47:59.830 --> 48:01.430]  There's a little contact on here.
[48:01.610 --> 48:02.350]  Always forgetting to tell people about the buttons.
[48:02.350 --> 48:04.990]  It's a double circle if you look on your badges.
[48:05.490 --> 48:05.930]  Yeah.
[48:06.430 --> 48:08.710]  We always thought that a button is a nice idea.
[48:08.710 --> 48:11.730]  But we simply don't have applications for it.
[48:11.770 --> 48:15.470]  So, in case you're interested to code nice stuff, you're warmly invited.
[48:15.610 --> 48:18.230]  So, basically what happens, you see this small circle.
[48:18.450 --> 48:21.430]  And you just have to short it with your thumb.
[48:21.730 --> 48:24.150]  So, not with your fingernail, but with your thumb.
[48:24.150 --> 48:27.110]  And then it will transmit a flag that is set.
[48:27.270 --> 48:29.250]  So, in our log files, it will appear.
[48:29.570 --> 48:30.550]  So, what will...
[48:30.550 --> 48:32.930]  We will immediately use it, I guess, in the visualization.
[48:33.290 --> 48:37.910]  So, in case you are in front of the main screen, you can press this button.
[48:38.250 --> 48:40.290]  And what I think what we'll do is to...
[48:40.290 --> 48:42.050]  That your avatar is increased.
[48:42.230 --> 48:43.430]  So, you know, okay, this...
[48:43.430 --> 48:46.470]  I press this and it immediately increases.
[48:46.610 --> 48:48.250]  So, you see where you are.
[48:48.550 --> 48:51.210]  But there are actually very cool things you can do.
[48:51.210 --> 48:54.650]  For example, you can download the log files later.
[48:54.950 --> 48:59.890]  And for example, you can decide to do bookmarking inside talks.
[49:00.190 --> 49:01.930]  So, for example, if you have a talk and...
[49:01.930 --> 49:02.850]  Oh, there is an interesting part.
[49:02.890 --> 49:06.090]  You just press on your button for a few seconds.
[49:06.290 --> 49:07.910]  So, this flag gets transmitted.
[49:08.110 --> 49:09.030]  It gets stored in the database.
[49:09.290 --> 49:13.390]  And later, you just need a tool that can...
[49:13.390 --> 49:18.310]  For this given time, can assign a bookmark to a video.
[49:18.310 --> 49:30.190]  So, I think for our next conference, what would be awesome is if we have a website, where you can just enter your tag ID and just jump to your bookmarks of the talks you attended.
[49:30.470 --> 49:32.330]  So, maybe this is a nice approach.
[49:32.490 --> 49:34.770]  But we can do all kinds of stuff.
[49:34.950 --> 49:38.130]  It's stored and you can do it for games, whatever.
[49:38.430 --> 49:39.910]  So, there is no particular...
[49:39.910 --> 49:49.190]  When we were talking about developing games, my suggestion was grenade mode, that the person with the most points at the time could hit the button and take points away from everyone else within range.
[49:50.950 --> 49:52.330]  I think that's it.
[49:52.410 --> 49:53.590]  We're kind of out of time.
[49:53.850 --> 49:54.070]  So, thank you.
[49:54.070 --> 49:55.430]  I wanted to add in one more thing.
[49:55.550 --> 49:58.110]  This is kind of a dynamic project through the duration of the cons.
[49:58.290 --> 50:01.370]  So, if you have an idea that you want to pitch in and like...
[50:01.370 --> 50:02.310]  We have all this cool data.
[50:02.390 --> 50:02.570]  We're taking code.
[50:02.790 --> 50:02.970]  Yeah.
[50:03.490 --> 50:06.050]  So, if you want to contribute, it's totally open.
[50:06.330 --> 50:08.170]  And like, please, programmers, come on.
[50:08.330 --> 50:09.290]  Let's do cool stuff.
[50:09.290 --> 50:11.070]  Or at least get us coffee on a regular basis.
[50:11.270 --> 50:11.790]  We could use it.
[50:12.930 --> 50:14.950]  So, for example, we can also...
[50:14.950 --> 50:20.450]  For example, one idea that was posted some months ago was, for example, to create a mood map.
[50:20.590 --> 50:33.930]  So, in case you like a particular talk or particular art installation, you just press this button multiple times and we can count in our system how often you press it and create some kind of live mood map around the building.
[50:33.930 --> 50:37.330]  So, you know where the most interesting parts are currently.
[50:37.330 --> 50:39.690]  So, but it needs to be implemented.
[50:39.970 --> 50:41.390]  So, we need people for this.
[50:41.690 --> 50:42.090]  Okay.
[50:42.490 --> 50:43.910]  So, these guys have to get back to work.
[50:44.350 --> 50:45.190]  So, join us.
[50:45.550 --> 50:46.950]  Because it's not running yet.
[50:47.810 --> 50:48.370]  Thank you.
[50:48.570 --> 50:49.410]  Thank you.