0:01 NARRATOR: Luminaries– talking to the brightest minds in tech.
0:05 MICHAEL DELL: And my hope is that we come together to share more than technology, expertise, and products, but that we share a vision of a future that is better than today, a vision of technology as the driver of human progress.
0:22 NARRATOR: Your hosts are Mark Schaefer and Douglas Carr.
0:26 [THEME MUSIC]
0:30 MARK SCHAEFER: Welcome, everyone, to another episode of Luminaries. This is Mark Schaefer with my super fun co-host Douglas Carr a.k.a. Dougy Baby. How you doing, Doug?
0:42 DOUGLAS KARR: I’m doing well. How are you, sir?
0:43 MARK SCHAEFER: Doug, I’ve never really asked you this before. Do you like sports? Are you a sporting kind of guy?
0:48 DOUGLAS KARR: I used to be a sports guy. Now I’m kind of a hiking guy.
0:52 MARK SCHAEFER: What are some of your favorite sports?
0:55 DOUGLAS KARR: I played football and broke my knee playing football and then I wrestled. And I loved wrestling.
1:01 MARK SCHAEFER: I love sports. I love all kinds of sports and here is one of my favorite sounds in the world of sports: Goal! We’re going to talk all about goals today. How was that, Daniel? I’m turning to my guest here. He has a look on his face that is absolutely priceless. So Daniel Marchwinski, how’d you like it? How’d I do?
1:32 DANIEL MARCHWINSKI: That was quite good for a nonprofessional.
1:36 MARK SCHAEFER: That’s awesome. I’ll take it. I’ll take it. So how does it go? How should it go?
1:43 DANIEL MARCHWINSKI: I’m not sure if I can do it. But I think the end was maybe– the voice was going out a little bit so you need a little more breath. You know the guys in Brazil or Mexico, they’re more like, go go go go go go go go goal! Go go go goal! So you might have done something like this because you get more breath into your lungs.
2:01 MARK SCHAEFER: I’m learning so much already. The people that are listening today are saying, what the hell is going on here? What are these guys talking about today? We are going to be talking about goals. Lots and lots and lots of goals, right? So you are an entrepreneur and I believe that your company is based in Germany. Is that right?
2:20 DANIEL MARCHWINSKI: Yep.
2:20 MARK SCHAEFER: Tell us about GoalControl.
2:23 DANIEL MARCHWINSKI: Yeah, of course. GoalControl is a company that provides goal line technology. That means, to say it very easily, we can help the referee to say is it a goal or not. For that, we have seven cameras around each goal with 400 frames per second that are filming the situation over there in the 16 meter box. And we combine all of these pictures together to a 3D application so that we can see if the ball crosses the goal line or not.
2:49 And we have an accuracy around 5 millimeters, which is quite good. And everything happens in real time so that in less than one second, we are able to help the referee to say it’s goal or it’s not a goal.
3:00 MARK SCHAEFER: And now your partner in this is a former professional soccer player, is that right?
3:08 DANIEL MARCHWINSKI: Exactly. We are owned by a soccer player. His name is Simon Rolfes. He is a national player from the Germany and he is quite well known over there in Europe. Not that much in America because you’re more on football than soccer, but if you see them you might like him.
3:24 MARK SCHAEFER: Well, I actually watched some of his YouTube highlights.
3:27 DOUGLAS KARR: I did too.
3:27 MARK SCHAEFER: And it was like, wow, he has some really great goals.
3:31 DANIEL MARCHWINSKI: Yeah, he’s a good guy. He’s a good guy. He was the team captain for a very long time from his team in the German Bundesliga Bayer Leverkusen. So it helps a lot and it’s good because he has all the contacts all over the world, to the football world, so that makes things for us much easier at the end of the day.
3:46 MARK SCHAEFER: So what was the spark of the idea for the company?
3:52 DANIEL MARCHWINSKI: At the end of the day, I think there are two milestones that brought us where we are. There was the Wembley goal, of course, where people the first time said, hey, it was a clear goal or not a goal. It’s still a discussion. And as a German, you know our opinion about that.
4:06 And then we had, in 2010, even a more important goal for Germany. It was Germany against Frank Lampard, or against England at the end of the day, in South Africa. And as a German, I have to say, it was a clear non-goal, just as the goal line was not straight. But no, that was a joke.
4:22 DANIEL MARCHWINSKI: It was clear in there. It was around 15 centimeters behind the goal line. And the referee didn’t see it. So there was a starting point where FIFA decided, hey, we have to do something. We have to help the referee to make the call, if it’s not a goal or if it’s a goal.
4:36 Think about sports getting much faster in the last 15 years. And also the audience grew up more, so more people are looking at it, and you have more eyes on the situation. So those are both the big milestones why we implemented that thing.
4:51 And the biggest thing for us, for goal control, was that we were the first company that has done goal line technology in the World Championships. So 2014, for the FIFA World Cup in Brazil, we were the official provider for goal line technology and we made two important decisions over there.
5:09 The first was France against Honduras. There was a first goal that ever was done with goal line technology. And we threw out the Italian guys, which is always nice for Germany.
5:19 MARK SCHAEFER: Love you, Italy. Love you, Italy. Don’t– we love you, Italy. So did I hear you correctly? Did the idea come from FIFA? They were saying, we need help?
5:30 DANIEL MARCHWINSKI: Yeah, the idea came from the football industry, of course. And at the same time we sought out a solution to do that. And the former owners of the company, the company that provided video based detection of objects, they decided.
5:44 For example, for a dinner plate, is there a bubble from the construction or from the building it in there? And they have put several cameras there to check every single plate. If there is something wrong with that, or still everything is OK. And from this idea today, if we can’t detect a mistake on a plate, we might be also able to see a ball.
6:09 It sounds so easy at the end of the day but it is quite difficult. It’s really, really difficult technology. There are only two providers worldwide that got the license to do that. And there was an application phase, where I have been around 15 guys who want to do it and only two got it. For such an easy thing, goal or non-goal.
6:32 DOUGLAS KARR: Well, let’s talk about such an easy thing, goal, non-goal. So you’ve got a round ball–
6:37 DANIEL MARCHWINSKI: Most of the time.
6:38 DOUGLAS KARR: –and then you’ve got a goal line. So at what point is it a goal?
6:43 DANIEL MARCHWINSKI: That’s a very good one. I had the discussion over there in our booth yesterday and also today. And some of them say, it’s absolutely clear to me when it’s a goal or not a goal.
6:52 And it’s not even in countries where football is very famous, people don’t know that the whole ball has to cross the line. Only then, it’s a goal. Like in tennis, for example, if there is a millimeter on the line, it’s still in.
7:06 For us, it’s exactly the opposite. It has to be completely over the line to make and score a goal. And this is, a lot of times, quite difficult to see, even if there is a fourth or fifth referee that is standing on the goal line.
7:18 DOUGLAS KARR: Right, and then let’s paint that picture. So you’ve got an entire stadium full of people. You can’t pause the game. You can’t sit and wait for the riots to erupt, right? You’ve got to be incredibly fast with the results with this.
7:34 DANIEL MARCHWINSKI: Exactly, and that is the biggest story, at the end of the day, that we have there. Like I said before, everything is done in real time. So in less than one second, the referee will get a signal on his watch when the ball crosses the goal line.
7:47 And if you compare it to a hot topic in Europe at the moment, where everybody’s talking about it, that is a VAR– that is a video assistant referee. This is where at least you can see, is it an offside? Did somebody hit the ball with his hand? Was it a fall or something like this?
8:03 This always interrupts the match because they have to check it. And our benefit, and our USB, is that we can say, OK, you just get, in less than one second, a signal on your watch. So it doesn’t interrupt the game. And that is the biggest thing for the audience in the stadium. You are celebrating a goal already and then the guy is coming with his fingers to his headset.
8:24 DANIEL MARCHWINSKI: –and he’s checking it, and after two minutes he says, no, it’s not a goal, sorry. This is not really nice for the audience in the stadium because they don’t know why and stuff like that.
8:33 And this is a big discussion over there in the European market at the moment. And our benefit is we don’t have the [INAUDIBLE]. We just give you the signal, it’s in or it’s not out. Easy decision, hard calculation behind that.
8:46 MARK SCHAEFER: So what does it look like? Tell us about how this is set up with the cameras.
8:52 DANIEL MARCHWINSKI: Yeah, we have on each goal seven cameras. They’re normally quite high in the stadium. So, if possible, and there is a catwalk, we will put them on the catwalk. So that you have a quite big angle on the ground so that not– the players can’t hide the view on the ball.
9:12 That is why we are going quite high over there. In general, to make a decision or a 3D position of the ball, we just need two cameras. But it’s always said, how should I say that– we have one big problem in football for our system. There are humans on the space so we have to be careful with that.
9:29 We don’t have all the time when we would have just two cameras, the possibility to see it. So we add seven cameras so that we have a 99.9% that we will definitely get two cameras that are seeing the ball in the right angle. And all of them are connected with high speed fiber connections so that we have a big calculation center where everything comes together.
9:52 And now that is a clue with that. We had a calculation center that was eight servers on each goal. So we had a van with two tons of technology in there just to do the calculation.
10:05 Now, with Dell, we found some solutions. How we can shrink that together and to get leaner. I know leaner is a big thing in ID and administration over there. So for us also.
10:16 So what we did, we came from CPU based calculation processes to GPU based calculation. And this makes it much easier and much, much faster for us. And it’s the first time that somebody implemented AI and machine learning also into football. So we are, in the past, we have just watched out for a round surface, like you say.
10:39 But, you know, there are some guys that might have lost their hair because they’re older or they drank something wrong, I don’t know. And they lost their hair, so it can be in one out of 100,000 cases that maybe the system will detect the ball or the bald guy as a ball. And he’s going into the goal, and it could give him the signal that it’s a goal.
11:00 So we decided to integrate artificial intelligence. So we are checking, is it a ball, yes or not? And only if it’s a ball, we will detect it, and we will follow this one. We’ll track it and see if it’s going on there.
11:11 And now, we are able, and that is maybe a very cool news for everybody. We have just one precision workstation there. A small workstation like a personal computer. And the system is running on this computer.
11:25 So we have three of them, we don’t need much space longer more in the stadium. And it is becoming a real plug and play solution now.
11:32 MARK SCHAEFER: Wow, that’s amazing. So what about configuration of the stadium or weather? I mean, do you run into problems with foggy or rainy conditions?
11:43 DANIEL MARCHWINSKI: Yeah, that’s a very good one. Of course, think about it, it’s a camera based system. It’s a video system.
11:51 So if you are sitting in the stadium and you can’t see anything because there is a firework in there, how should a camera see it? It’s exactly the same. So, normally, snow, rain, wind, is not interfering our system. We don’t have the impact on that one.
12:07 But if there is some dust in the stadium, or fireworks will be there, this might have an impact on that and we have to say, sorry, the system is not working on that one at the moment.
12:17 DOUGLAS KARR: Interesting. Now, it’s interesting. You guys went from a camera based system instead of a ball based system. Was that to keep the ball perfect? You didn’t want to touch the–
12:28 DANIEL MARCHWINSKI: Yeah, at the end of the day, if you put something into the ball, it might interfere with the ball. So it might not go straight and stuff like that and people don’t like that.
12:38 Or what happened if somebody is throwing in the wrong ball. So it’s not longer more possible to detect it. And what we say, and what also the FIFA say, is that it’s super important that the character of the game and even of the– how to say that? –of the environment over there.
12:55 So the pitch with the goals and everything will not be touched. And the only possibility to do that is just to have cameras as far away as possible. So we also had ideas about integrating cameras into the goal frame.
13:09 But think about it, somebody is hitting the goal frame with a ball with 150 kilometers per hour. I’m not an expert but I would say, it will definitely make something broken in the frame. So it might be the mirror or maybe the camera is broken.
13:22 MARK SCHAEFER: Even the vibration.
13:23 DANIEL MARCHWINSKI: Even the vibration. We see that even if somebody is jumping on the stands we see that this infects the camera. So what we have implemented over there is that we have like a filter that is filtering out the vibrations over there dramatically.
13:40 MARK SCHAEFER: So it’s an interesting point because you’re talking about, really the purity of the pitch and the environment and it is a beautiful game. It’s such a human game.
13:53 And just like in America, there’s a lot of passion that goes along with this sport. Every time there’s a technological intervention in one of our sporting events, there’s an outcry from the purists. You’re taking away part of the game.
14:08 So are you facing something like that? And is it well accepted now?
14:14 DANIEL MARCHWINSKI: Absolutely, we have exactly the same problems. They think you destroyed the game. Sorry to say, that’s just not true. We make it just a little bit more fair.
14:23 And, in our case, I think it was not so big, the problem over there. Because it’s not stopping the game at all. It’s just one second, we’re getting the signal.
14:37 But in general, and in football, implementing technology is quite difficult. The FIFA, quite old guys are sitting there, it is like, a lot of people think about that and make decision. And they are maybe not so friendly against innovation. So yes, it is quite difficult. We’re facing a lot of problems over there.
14:57 MARK SCHAEFER: How did you even prove it? Because this is something completely new. It’s never been done before.
15:04 So what did FIFA say? You’ve got to– what? Did you have someone just kick goals for a week through the goal? I mean, how do you prove this thing?
15:12 DANIEL MARCHWINSKI: The system. Yeah, that’s a good one. In general, everybody who wants to provide technology, in the world of sports or the world of football, has to get a FIFA license for that.
15:22 And you have a license agreement where they say what has to be done so that you get it. This is done by an external testing institute. And they will do several tests on it.
15:35 They will do, like you say, some shots on the goal, maybe not a week, but a few hours. They will run with the ball in their hands into the goal. Several things that they will test over there to see if the system is working or not.
15:46 And only if you get this license, you are allowed to install the technology in a stadium. And then you have to do a so-called fit test, what is a final installation test, where you’re getting a certificate for each stadium, for each ball, for one year.
16:01 It’s quite a big investment also to certify every single stadium that’s there in your system. But this gives you the comfort or the freedom that the system is working quite well over there.
16:15 But even then, there might be going something wrong because they are just humans using the system. So you have to adjust the brightness, the color, and stuff like that. But yeah, there are some tests.
16:26 MARK SCHAEFER: You really need a very skilled operator at every single game.
16:30 DANIEL MARCHWINSKI: Yes. We need that but we are working now on something more like, that’s my vision. And that’s maybe an important thing.
16:39 My vision is that we’re coming to a stage where there’s a green and a red button. OK, that is where I want to come. So everything will be referenced automatically.
16:48 Even a dumb person can just go there, push the green or red button, at least he can see colors, yeah? So maybe we should also say “play” and “stop” on it or something like this? But then there’s a problem he can’t read. I forgot that. Maybe we should put some symbols on it.
17:05 DANIEL MARCHWINSKI: But we can put something on there, and it’s just start or stop, plug and play. Nobody has to operate it longer more. This is where we want to come.
17:13 I don’t know when we will come there, but we will definitely come there. First step will to use it remote. And then maybe someday we will come to the green and red button.
17:21 DOUGLAS KARR: The solution is described as 4D. So not–
17:26 DANIEL MARCHWINSKI: I have no idea why. It’s just marketing.
17:28 DANIEL MARCHWINSKI: No. It’s quite easy, because we’re using the XYZ coordinates and time. So that is why we call it 4D. So we can say to any single moment where in 3D room the ball is.
17:41 So that’s why it’s 4D. But if I’m honest, hand on my heart, I don’t like the name here. It’s not cool enough. We should call it Iron Man 3 or 4, or something like this. So that people love it.
17:54 MARK SCHAEFER: Mark, strong name. The Mark 4.
17:58 DANIEL MARCHWINSKI: I love it. You know, the Mark III, the camera model from–
18:02 MARK SCHAEFER: I’ll licence it to you. We’ll talk about that after the podcast. So I’m a big baseball fan and there’s a lot of people in baseball that would love to have something like you have for baseball for the strike zone. Are you starting to think about other types of sports?
18:21 DANIEL MARCHWINSKI: Yes, of course. So we have already two years ago, the possibility to present our system to the NBA. Where we trekked the ball over the full field and see how it’s going on there.
18:32 And it’s quite difficult because it’s a brown ball and some of the players might be a little bit brown from their color. So it was a tough thing to do that but we were able to track the ball over the whole field over there.
18:42 And definitely, when we have implemented now our new technology with AI and machine learning in there, next steps will definitely be looking at basketball, baseball also.
18:53 I learned today, baseball is the biggest sport in America and the most famous one. I always thought it was basketball, but it’s baseball. And wow, we are more than happy to help these guys over there. So they should just come to us and we start the conversation over there. So if you hear us, we are happy. I will give you my number later on.
19:13 MARK SCHAEFER: Here you talked about the tipping point for you was the Wembley goal? So in a baseball game, we have a Wembley goal every single day.
19:24 Even the announcers, where it’s like, that was not a strike! How could they have called that a strike? And the players are stomping up and down. That happens every game, every day, every team plays 162 games. So there is an infinite opportunity for you, sir.
19:41 DANIEL MARCHWINSKI: We go for that. Let’s start this conversation later on.
19:45 DOUGLAS KARR: One last question about the technology. So you said that there are seven cameras that are put within the stadium that are sitting, obviously, a great distance. It’s a fiber connection to communicate with them. What kind of resolutions and data is getting passed?
20:00 DANIEL MARCHWINSKI: It’s a quite high amount of data that are going there. I can’t tell you how much it is completely. I can tell you if you would save all the video data from one match without detecting the ball on it, just the video data, it would be around 50 terabytes per half. So that means 100 terabytes of data that are going around there on each match.
20:21 DOUGLAS KARR: Wow.
20:22 DANIEL MARCHWINSKI: So we had a discussion with the Dell guys already if they have storage solutions for that. And yeah, we’ll go ahead with that. For the moment, we are just saving the position, or let’s say, the time where there was action and a goal was coming or a wide, close to the goal, situation that will be a non-goal.
20:40 Everything else, we’re not storing at the moment because we can’t. But, think about it, we have 400 frames per second on each camera, seven cameras on each goal, 14 cameras in total. It’s a huge amount of data that is going around there. And everything is in real time.
20:54 DOUGLAS KARR: Right, and the decision is in–
20:56 DANIEL MARCHWINSKI: Real time.
20:58 MARK SCHAEFER: And you mentioned that FIFA really has– there’s lots of competitors trying to get in this. FIFA has only approved two? So you’re really onto something here. So what are you doing to drive your company to stay ahead? To make sure that you are the preferred solution?
21:18 DANIEL MARCHWINSKI: It’s just the easy word: innovation. And we want to innovate and we want to be the best one in this field. And for this, we need partners, good partners, like Dell, for example.
21:29 I told already the whole day, the story that we need a precision workstation but with a very special configuration that nobody else would ever use. It’s still everyday components but how they are combined, it’s not normal.
21:44 So what has Dell done? They implemented our computer that we need– nobody else in the world would need it– in their normal program so that we can order it to any time. So that’s amazing. That’s really good. So innovations are super important for us.
21:57 And to have a look at the newest technology like AI and the machine learning. That is why we want to keep ahead. And, of course, what is an important thing for us? What we’re working on is to give a cash back on it.
22:12 Let’s say it this way, at the moment it just costs money to a league or to a stadium or a club, or whatever. If he is able to earn money with that, it becomes more interesting for them.
22:22 MARK SCHAEFER: Oh, that is interesting. So what is your monetization? Is it the installation? Is it a license? Is it per game? Or how does that work for you?
22:32 DANIEL MARCHWINSKI: Oh, what we are normally doing, we are selling our system, what is a one time shot. And then we have yearly license fee on it. And, of course, the cost for installation and operating the system.
23:12 MARK SCHAEFER: So, Daniel, this has been absolutely fascinating. So if our listeners want to learn more about you and what you’re doing with GoalConrol, what’s the best way to find you on the web and follow your progress?
23:27 DANIEL MARCHWINSKI: Yeah, they can go just to our home page and have a look at this one. We are also quite active on Facebook and Twitter. We starting now Twitter. But we will go ahead with it. But have a look at our home page and they will see some videos, nice materials over there, how it works and what we did before already.
23:45 MARK SCHAEFER: Awesome. Thank you so much, it’s been really, really interesting and a super fun interview. And thanks to all of you for listening to Luminaries. We appreciate every one of you. Thank you for your loyalty. You’ve made us one of the top 1% of all business podcasts, so we really appreciate that.
24:05 This is Mark Schaefer, and on behalf of Doug Carr, my co-host, we thank you so much. We’ll see you next time on Luminaries.
24:12 [THEME MUSIC]
24:14 NARRATOR: Luminaries– talking to the brightest minds in tech. A podcast series from Dell Technologies.