[
  {
    "start": 3.82,
    "end": 16.29,
    "text": "From 5G networks to aerospace systems, from artificial intelligence to quantum technologies, from behind-the-scenes of innovations to real-world applications.",
    "speaker_id": null
  },
  {
    "start": 17.23,
    "end": 20.09,
    "text": "Welcome to INNOVATIONS UNPLUGGED",
    "speaker_id": null
  },
  {
    "start": 20.85,
    "end": 23.65,
    "text": "The Rohde & Schwarz Technology Podcast.",
    "speaker_id": null
  },
  {
    "start": 25.89,
    "end": 27.19,
    "text": "I am Markus Haller.",
    "speaker_id": null
  },
  {
    "start": 27.53,
    "end": 32.03,
    "text": "And I'm Inga Müller-Siedentopp and we're both happy to host this show.",
    "speaker_id": null
  },
  {
    "start": 32.549,
    "end": 40.31,
    "text": "Today we're pulling back the curtain on a tech that's older than you might think, but still speeding into the future.",
    "speaker_id": null
  },
  {
    "start": 40.91,
    "end": 41.19,
    "text": "Radar.",
    "speaker_id": null
  },
  {
    "start": 41.81,
    "end": 47.91,
    "text": "To be more precise, today we talk about hyper-realistic radar simulation.",
    "speaker_id": null
  },
  {
    "start": 48.83,
    "end": 54.03,
    "text": "We'll take a closer look at how radar already powers everyday digital life.",
    "speaker_id": null
  },
  {
    "start": 54.55,
    "end": 57.31,
    "text": "We'll explore hyper-realistic radar simulation.",
    "speaker_id": null
  },
  {
    "start": 57.629,
    "end": 60.93,
    "text": "And how artificial intelligence enters the game here.",
    "speaker_id": null
  },
  {
    "start": 61.73,
    "end": 64.67,
    "text": "and we'll find out why it matters to all of us.",
    "speaker_id": null
  },
  {
    "start": 65.71,
    "end": 69.229,
    "text": "Joining us are Marcel Hoffmann and Michael Stelzig.",
    "speaker_id": null
  },
  {
    "start": 69.509,
    "end": 82.57,
    "text": "They are the co-founders and managing directors of fiveD, an innovation leader in hyper-realistic radar simulations, Rohde & Schwarz joint forces with.",
    "speaker_id": null
  },
  {
    "start": 83.229,
    "end": 85.31,
    "text": "Welcome Marcel, welcome Michael.",
    "speaker_id": null
  },
  {
    "start": 86.31,
    "end": 87.39,
    "text": "Thank you very much for having us.",
    "speaker_id": null
  },
  {
    "start": 88.09,
    "end": 88.63,
    "text": "Thank you as well.",
    "speaker_id": null
  },
  {
    "start": 89.49,
    "end": 90.009,
    "text": "Happy to be here.",
    "speaker_id": null
  },
  {
    "start": 91.17,
    "end": 95.47,
    "text": "Michael, could you tell us where radar helps us every day?",
    "speaker_id": null
  },
  {
    "start": 96.07,
    "end": 99.89,
    "text": "It's basically a technology that touches you in more places than you would expect:",
    "speaker_id": null
  },
  {
    "start": 100.289,
    "end": 105.07,
    "text": "Starts at the automatic toilet flushing because there's a radar sensor behind the tiles for example,",
    "speaker_id": null
  },
  {
    "start": 105.11,
    "end": 108.33,
    "text": "that's a really simple radar system and it ends in.",
    "speaker_id": null
  },
  {
    "start": 108.45,
    "end": 118.13,
    "text": "Super complex systems are used for example for microwave body scanners at the airport where a radar system searches for hazardous objects on your body",
    "speaker_id": null
  },
  {
    "start": 118.75,
    "end": 121.99,
    "text": "or you have it on airports for surveillance of the air.",
    "speaker_id": null
  },
  {
    "start": 122.11,
    "end": 125.05,
    "text": "You have it in aerospace defense applications,",
    "speaker_id": null
  },
  {
    "start": 125.289,
    "end": 131.99,
    "text": "cars use it for autonomous driving so that the car is able to see it's the environment for smart home applications.",
    "speaker_id": null
  },
  {
    "start": 132.03,
    "end": 135.21,
    "text": "So there are so many touch points with radar that you don't even know of.",
    "speaker_id": null
  },
  {
    "start": 135.39,
    "end": 136.53,
    "text": "Okay, that's fascinating.",
    "speaker_id": null
  },
  {
    "start": 136.81,
    "end": 143.329,
    "text": "And I did not expect it to be that close to me - talking about automotive.",
    "speaker_id": null
  },
  {
    "start": 144.43,
    "end": 147.99,
    "text": "We don't have to go to autonomous driving yet.",
    "speaker_id": null
  },
  {
    "start": 148.23,
    "end": 152.89,
    "text": "I think automatic cruise control, this also relies on radar sensors, right?",
    "speaker_id": null
  },
  {
    "start": 153.41,
    "end": 154.329,
    "text": "Yes, that is correct.",
    "speaker_id": null
  },
  {
    "start": 154.63,
    "end": 164.549,
    "text": "Measuring the distance and the velocities of other objects in the driving scenario, so other cars, pedestrians, but also buildings, guardrails whatsoever.",
    "speaker_id": null
  },
  {
    "start": 164.73,
    "end": 169.41,
    "text": "The radar is a really good technology to make the car able to see its environment.",
    "speaker_id": null
  },
  {
    "start": 170.31,
    "end": 174.91,
    "text": "Now Marcel, fiveD is not building hardware.",
    "speaker_id": null
  },
  {
    "start": 174.99,
    "end": 182.53,
    "text": "You are not developing radar systems yourself, but what you are doing is you create radar simulation software.",
    "speaker_id": null
  },
  {
    "start": 183.25,
    "end": 185.63,
    "text": "What exactly does fiveD do?",
    "speaker_id": null
  },
  {
    "start": 186.41,
    "end": 190.73,
    "text": "We have developed a simulation solution that helps develop radar systems.",
    "speaker_id": null
  },
  {
    "start": 191.73,
    "end": 203.37,
    "text": "The simulation allows you to test things that are extremely difficult and expensive and time-consuming to smoothen and improve and streamline the real hardware development process.",
    "speaker_id": null
  },
  {
    "start": 203.47,
    "end": 204.75,
    "text": "But it doesn't stop at the hardware.",
    "speaker_id": null
  },
  {
    "start": 205.21,
    "end": 206.31,
    "text": "It goes over to software.",
    "speaker_id": null
  },
  {
    "start": 206.37,
    "end": 211.25,
    "text": "It goes all the way to find out if a sensor concept can work in real-world applications.",
    "speaker_id": null
  },
  {
    "start": 211.69,
    "end": 212.85,
    "text": "Can you walk us through?",
    "speaker_id": null
  },
  {
    "start": 213.049,
    "end": 215.13,
    "text": "how does it really help?",
    "speaker_id": null
  },
  {
    "start": 215.25,
    "end": 225.31,
    "text": "Like what pain points do engineers face when they develop radar systems, radar technology for our cars?",
    "speaker_id": null
  },
  {
    "start": 225.79,
    "end": 232.69,
    "text": "I can give you a very specific or my own personal pain point that I had or that we had during our time at the university.",
    "speaker_id": null
  },
  {
    "start": 233.15,
    "end": 244.01,
    "text": "The process for a PhD student for example involves developing a sensor, developing a sensor concept, setting up the algorithms and software for it and bringing it to the street, which is impossible within the timeframe of a PhD student.",
    "speaker_id": null
  },
  {
    "start": 244.87,
    "end": 255.63,
    "text": "So we came up with the idea of we need a tool to simulate that, to simulate realistic street scenarios, to get the application into the office because we can't bring it to the street.",
    "speaker_id": null
  },
  {
    "start": 256.53,
    "end": 265.07,
    "text": "And this was our personal pain point, which led us to developing the simulator because we didn't find a solution on the market that could solve this problem.",
    "speaker_id": null
  },
  {
    "start": 265.77,
    "end": 278.71,
    "text": "But this is a similar problem that big companies and established tier ones and tier twos and OEMs and automotive also have because prototyping no matter how is expensive and time consuming and you don't want to be lacking behind on that.",
    "speaker_id": null
  },
  {
    "start": 280.33,
    "end": 285.85,
    "text": "You don't want to spend a year or two setting up the sensor and then a year or two prototyping and then gathering data for a year and then",
    "speaker_id": null
  },
  {
    "start": 286.47,
    "end": 291.21,
    "text": "you don't want to find out, oh, I should start back at the prototyping stage because something wasn't ideal.",
    "speaker_id": null
  },
  {
    "start": 291.47,
    "end": 296.79,
    "text": "It's even more complicated than that because the sensor in the car is typically not a standalone system.",
    "speaker_id": null
  },
  {
    "start": 296.89,
    "end": 300.23,
    "text": "It goes hand in hand with other systems and it has to be integrated in the car.",
    "speaker_id": null
  },
  {
    "start": 300.65,
    "end": 308.11,
    "text": "The baseline is real world testing and real world development is expensive and takes time and we can speed this up with simulation.",
    "speaker_id": null
  },
  {
    "start": 310.15,
    "end": 315.15,
    "text": "To picture this automotive engineer pain points pretty good right now.",
    "speaker_id": null
  },
  {
    "start": 316.55,
    "end": 322.71,
    "text": "Drones are another example that are becoming increasingly relevant today.",
    "speaker_id": null
  },
  {
    "start": 322.91,
    "end": 326.83,
    "text": "I think that is not to discuss, it's just fact.",
    "speaker_id": null
  },
  {
    "start": 327.75,
    "end": 329.31,
    "text": "But they are very small.",
    "speaker_id": null
  },
  {
    "start": 329.83,
    "end": 332.63,
    "text": "They are diverse and they are hard to detect.",
    "speaker_id": null
  },
  {
    "start": 333.01,
    "end": 335.49,
    "text": "Let's put it to a really specific example.",
    "speaker_id": null
  },
  {
    "start": 335.59,
    "end": 341.31,
    "text": "We had a few weeks ago this problem that the operations at the Munich airport stopped because of drone",
    "speaker_id": null
  },
  {
    "start": 342.05,
    "end": 342.25,
    "text": "sight.",
    "speaker_id": null
  },
  {
    "start": 343.47,
    "end": 353.71,
    "text": "Next step could be, hey, let's evaluate a sensor to set up how to protect the perimeter of the overall airport and evaluate where should the sensors and which sensors where should they be.",
    "speaker_id": null
  },
  {
    "start": 354.23,
    "end": 360.79,
    "text": "So we have the option now to make real world testing put the sensors there and then test it.",
    "speaker_id": null
  },
  {
    "start": 360.85,
    "end": 368.57,
    "text": "but for this you have to fly with the drones from all sorts of directions at different heights different types of drones and you can guess it already.",
    "speaker_id": null
  },
  {
    "start": 369.05,
    "end": 375.41,
    "text": "You have to do that during operations which is not possible so you don't have to the time or the options to do those tests.",
    "speaker_id": null
  },
  {
    "start": 375.75,
    "end": 385.23,
    "text": "Then you could say okay Munich Airport you can't fly at night we do it at night yes but then all of the other other sensor technologies can't be tested because cameras don't work in the night.",
    "speaker_id": null
  },
  {
    "start": 385.81,
    "end": 392.55,
    "text": "And that's kind of where also stimulation come into play because there you just don't have the options to test them in real life.",
    "speaker_id": null
  },
  {
    "start": 393.09,
    "end": 397.33,
    "text": "That's for the civil applications, for the defense application that gets even",
    "speaker_id": null
  },
  {
    "start": 397.37,
    "end": 397.63,
    "text": "worse.",
    "speaker_id": null
  },
  {
    "start": 400.11,
    "end": 409.29,
    "text": "We do want to be faster when we develop radars or actually companies want to be faster, PhD students want it to be faster.",
    "speaker_id": null
  },
  {
    "start": 409.51,
    "end": 412.83,
    "text": "And also it's about the other specs around prototyping.",
    "speaker_id": null
  },
  {
    "start": 412.93,
    "end": 414.65,
    "text": "You need sensors.",
    "speaker_id": null
  },
  {
    "start": 415.43,
    "end": 419.21,
    "text": "and test them in the real world, which is not an option every time.",
    "speaker_id": null
  },
  {
    "start": 419.41,
    "end": 425.95,
    "text": "So, when engineers listen, they might think, OK, but simulation itself is not so new for us.",
    "speaker_id": null
  },
  {
    "start": 426.61,
    "end": 430.25,
    "text": "So, I would say there are different levels of simulation.",
    "speaker_id": null
  },
  {
    "start": 430.73,
    "end": 437.93,
    "text": "So, when we look at the development timeline of a radar, for example, it starts really early stage.",
    "speaker_id": null
  },
  {
    "start": 438.13,
    "end": 440.53,
    "text": "You develop, for example, a radar chip.",
    "speaker_id": null
  },
  {
    "start": 440.87,
    "end": 443.37,
    "text": "There are simulation tools for simulating chips.",
    "speaker_id": null
  },
  {
    "start": 443.93,
    "end": 444.61,
    "text": "You have the antenna.",
    "speaker_id": null
  },
  {
    "start": 445.15,
    "end": 446.95,
    "text": "There are tools to simulate the antenna.",
    "speaker_id": null
  },
  {
    "start": 447.35,
    "end": 454.61,
    "text": "They typically rely on solving the Maxwell equations in a three-dimensional volume.",
    "speaker_id": null
  },
  {
    "start": 455.11,
    "end": 457.55,
    "text": "That is extremely computationally expensive.",
    "speaker_id": null
  },
  {
    "start": 458.17,
    "end": 469.15,
    "text": "So, when I now look for example at the automotive case where a car needs to see like four hundred meters you have to solve the equations in a like twenty times per wavelength.",
    "speaker_id": null
  },
  {
    "start": 469.63,
    "end": 472.31,
    "text": "Roughly, we're talking about three millimeter wavelength.",
    "speaker_id": null
  },
  {
    "start": 472.37,
    "end": 474.61,
    "text": "Now you can imagine how large the volume is.",
    "speaker_id": null
  },
  {
    "start": 475.13,
    "end": 478.67,
    "text": "How many times do you have to solve these really complex equations?",
    "speaker_id": null
  },
  {
    "start": 478.93,
    "end": 478.93,
    "text": "A",
    "speaker_id": null
  },
  {
    "start": 479.07,
    "end": 481.25,
    "text": "lot of computational effort is needed.",
    "speaker_id": null
  },
  {
    "start": 481.29,
    "end": 481.97,
    "text": "It is not working.",
    "speaker_id": null
  },
  {
    "start": 482.43,
    "end": 488.55,
    "text": "We are not at the point that we have enough computational power and we won't be there for quite some years to solve this with these approaches.",
    "speaker_id": null
  },
  {
    "start": 488.59,
    "end": 489.63,
    "text": "Just make it an exclamation mark.",
    "speaker_id": null
  },
  {
    "start": 490.39,
    "end": 491.79,
    "text": "We are just not able to, right?",
    "speaker_id": null
  },
  {
    "start": 491.83,
    "end": 491.93,
    "text": "Yes.",
    "speaker_id": null
  },
  {
    "start": 493.07,
    "end": 493.81,
    "text": "That's the one point.",
    "speaker_id": null
  },
  {
    "start": 493.85,
    "end": 497.25,
    "text": "Then at the other end of the spectrum, we have simulations for functions.",
    "speaker_id": null
  },
  {
    "start": 497.45,
    "end": 506.19,
    "text": "So we have radar camera, LiDAR put together for example, for fusion algorithm, then then it's used for path planning in autonomous driving.",
    "speaker_id": null
  },
  {
    "start": 506.81,
    "end": 508.65,
    "text": "That's really high level simulation.",
    "speaker_id": null
  },
  {
    "start": 508.71,
    "end": 512.189,
    "text": "It's optimized for real time and there the realism",
    "speaker_id": null
  },
  {
    "start": 512.75,
    "end": 514.77,
    "text": "of the radar signal is not given.",
    "speaker_id": null
  },
  {
    "start": 514.85,
    "end": 523.25,
    "text": "So, to have a proper tool for evaluating what's the influence by changing small bits on my radar is not reflected in the final result.",
    "speaker_id": null
  },
  {
    "start": 523.75,
    "end": 535.689,
    "text": "And that's exactly what we want to do to fill this gap between those super precise simulations that we can't do on a larger volume and the larger volume simulation that are not precise enough.",
    "speaker_id": null
  },
  {
    "start": 536.05,
    "end": 537.93,
    "text": "And we are exactly fitting in this gap.",
    "speaker_id": null
  },
  {
    "start": 538.27,
    "end": 544.67,
    "text": "Inga, regarding your question, I would like to plant a thought that is relevant when talking about radar.",
    "speaker_id": null
  },
  {
    "start": 545.43,
    "end": 552.069,
    "text": "For radar we have to make sure that we think from a radar perspective and radar does not work as visible light.",
    "speaker_id": null
  },
  {
    "start": 552.63,
    "end": 554.79,
    "text": "Trees are green and hills are, I don't know, brown.",
    "speaker_id": null
  },
  {
    "start": 556.469,
    "end": 565.63,
    "text": "Radar sees with a significantly reduced resolution, so shapes and contours are not as relevant as for visual light.",
    "speaker_id": null
  },
  {
    "start": 566.73,
    "end": 568.93,
    "text": "Instead we're interested in the reflectivity of things.",
    "speaker_id": null
  },
  {
    "start": 569.03,
    "end": 571.069,
    "text": "How strong does a tree or a hill reflect?",
    "speaker_id": null
  },
  {
    "start": 571.67,
    "end": 577.449,
    "text": "And we're also interested and that's one very interesting aspect that brings in new opportunities.",
    "speaker_id": null
  },
  {
    "start": 577.89,
    "end": 580.569,
    "text": "We're interested in velocity and motion components.",
    "speaker_id": null
  },
  {
    "start": 581.089,
    "end": 584.569,
    "text": "Radar is extremely good at finely detecting motion.",
    "speaker_id": null
  },
  {
    "start": 585.469,
    "end": 589.77,
    "text": "For example vibration patterns of machines or cars moving in the city.",
    "speaker_id": null
  },
  {
    "start": 591.41,
    "end": 598.89,
    "text": "So, that radar data looks completely different than what we would think of if we think of a hill with trees and these types of things.",
    "speaker_id": null
  },
  {
    "start": 599.39,
    "end": 611.43,
    "text": "You can maybe compare it to an ultrasonic image at a doctor's office where the practitioner tells you, what you see here is something wrong in your kidney and you just say, ah, okay, but you really don't see it.",
    "speaker_id": null
  },
  {
    "start": 611.89,
    "end": 615.27,
    "text": "But that analogy works quite well with radar.",
    "speaker_id": null
  },
  {
    "start": 615.589,
    "end": 616.91,
    "text": "It's nothing we would interpret.",
    "speaker_id": null
  },
  {
    "start": 617.329,
    "end": 623.65,
    "text": "as we would interpret a visual image, we detect reflectivity and we detect motion and these types of things.",
    "speaker_id": null
  },
  {
    "start": 624.37,
    "end": 627.03,
    "text": "And just to emphasize this, maybe a little story from myself.",
    "speaker_id": null
  },
  {
    "start": 627.11,
    "end": 635.93,
    "text": "When I first saw what data you are dealing with, what diagrams, and I, as a physicist, I am used to simplify things.",
    "speaker_id": null
  },
  {
    "start": 636.11,
    "end": 652.79,
    "text": "And when you showed me, this is a model of a pedestrian walking towards and away from a radar sensor, I was actually expecting to see one dot in this diagram, but what I was seeing is a line, a horizontal line.",
    "speaker_id": null
  },
  {
    "start": 655.069,
    "end": 657.85,
    "text": "Where does this several dots come from?",
    "speaker_id": null
  },
  {
    "start": 658.089,
    "end": 672.77,
    "text": "And the answer is the radar does not see the pedestrian as a point, but it sees the torso of a pedestrian moving at a certain velocity, the hand swinging in front, the other hand swinging back, and all this data is doing a line.",
    "speaker_id": null
  },
  {
    "start": 673.31,
    "end": 676.31,
    "text": "And I really had to adjust my thinking when seeing this data.",
    "speaker_id": null
  },
  {
    "start": 676.969,
    "end": 682.49,
    "text": "The point that Markus just raised is extremely important and that's where simulation plays an extremely huge role.",
    "speaker_id": null
  },
  {
    "start": 682.99,
    "end": 690.089,
    "text": "Because you just mentioned that there is different motion components like the swinging arms and the legs and the torso and these types of things.",
    "speaker_id": null
  },
  {
    "start": 690.89,
    "end": 698.01,
    "text": "And as you've just said this is nothing our brain is expecting to see and this is nothing our brain would be very good at classifying.",
    "speaker_id": null
  },
  {
    "start": 698.31,
    "end": 703.91,
    "text": "A car needs to classify the targets and the objects it sees in its surrounding to find out if it should stop or not.",
    "speaker_id": null
  },
  {
    "start": 704.35,
    "end": 709.51,
    "text": "And this is something that we are not fully used to doing because we're used to seeing visual images.",
    "speaker_id": null
  },
  {
    "start": 710.449,
    "end": 715.75,
    "text": "But the motion profile that you just described, Markus, is what we call a micro-doppler signature.",
    "speaker_id": null
  },
  {
    "start": 716.37,
    "end": 723.189,
    "text": "It's an inherent signature like a fingerprint that would give the specific motion components that a moving object has.",
    "speaker_id": null
  },
  {
    "start": 723.709,
    "end": 728.069,
    "text": "And this fingerprint can be extremely reliable, classified with, for example, AI.",
    "speaker_id": null
  },
  {
    "start": 728.87,
    "end": 736.53,
    "text": "And we need to make sure that this AI gets all the training information that it needs to be able to learn and understand what it just saw.",
    "speaker_id": null
  },
  {
    "start": 737.469,
    "end": 744.11,
    "text": "And again, this is something where simulation comes in, because simulation data solves the problem of training AI.",
    "speaker_id": null
  },
  {
    "start": 744.23,
    "end": 759.29,
    "text": "It brings in all the information that you need to give fully labeled, is what we call it, fully explained data sets to an AI to tell the AI what it is currently seeing, to detect patterns out of this, to detect pedestrians reliably.",
    "speaker_id": null
  },
  {
    "start": 761.589,
    "end": 762.61,
    "text": "There was a perfect bridge.",
    "speaker_id": null
  },
  {
    "start": 762.75,
    "end": 773.73,
    "text": "Actually because we were wondering, you claim very prominently on your website from raw data to next level radar AI.",
    "speaker_id": null
  },
  {
    "start": 773.85,
    "end": 781.11,
    "text": "And I think we need to take a step back here because as far as we understood it, your software itself is not AI driven.",
    "speaker_id": null
  },
  {
    "start": 781.329,
    "end": 788.27,
    "text": "Very good point and I'm really thankful that you mentioned the point that we're not AI driven, but we're software that can enable AI.",
    "speaker_id": null
  },
  {
    "start": 788.99,
    "end": 792.81,
    "text": "And if you look at how AI works, it's always the processor.",
    "speaker_id": null
  },
  {
    "start": 792.91,
    "end": 798.17,
    "text": "The first, you need large label data sets to feed into the AI training.",
    "speaker_id": null
  },
  {
    "start": 798.25,
    "end": 801.05,
    "text": "So, basically it's the same thing as when you raise a child.",
    "speaker_id": null
  },
  {
    "start": 801.449,
    "end": 806.209,
    "text": "You tell the child, super many times, that's a tree, that's a car, that's a house.",
    "speaker_id": null
  },
  {
    "start": 806.89,
    "end": 807.01,
    "text": "Yeah.",
    "speaker_id": null
  },
  {
    "start": 807.469,
    "end": 812.589,
    "text": "And at some point the child learns, okay, I get it myself and look, mom, look that, that's a tree, right?",
    "speaker_id": null
  },
  {
    "start": 812.75,
    "end": 814.209,
    "text": "And you say, no, that's a bush.",
    "speaker_id": null
  },
  {
    "start": 814.73,
    "end": 817.17,
    "text": "But at some point it will learn it.",
    "speaker_id": null
  },
  {
    "start": 817.51,
    "end": 820.65,
    "text": "And for this, it just needs to see it over and over and over again.",
    "speaker_id": null
  },
  {
    "start": 820.709,
    "end": 824.05,
    "text": "And it's the same thing if you need to train and radar AI.",
    "speaker_id": null
  },
  {
    "start": 825.27,
    "end": 830.67,
    "text": "So, now for radar, as you already mentioned, if you look at the data, you can't really see what it is.",
    "speaker_id": null
  },
  {
    "start": 831.11,
    "end": 835.41,
    "text": "So, to label this and then to train the machine is a super hard task.",
    "speaker_id": null
  },
  {
    "start": 836.13,
    "end": 838.75,
    "text": "And the cool thing with simulation now, you know it already.",
    "speaker_id": null
  },
  {
    "start": 839.469,
    "end": 843.89,
    "text": "Because we have built this 3D world, we simulate how the radar sees this world.",
    "speaker_id": null
  },
  {
    "start": 844.35,
    "end": 848.89,
    "text": "And in the signal itself, we already have the information what the radar has seen.",
    "speaker_id": null
  },
  {
    "start": 849.329,
    "end": 853.65,
    "text": "And therefore, inherently, we have the perfectly labeled data available.",
    "speaker_id": null
  },
  {
    "start": 854.03,
    "end": 860.01,
    "text": "So now, when you set up the simulation pipeline of this realistic simulation, just click Run.",
    "speaker_id": null
  },
  {
    "start": 860.63,
    "end": 861.829,
    "text": "You have a large data set.",
    "speaker_id": null
  },
  {
    "start": 861.969,
    "end": 863.67,
    "text": "You can vary it by a click of a mouse.",
    "speaker_id": null
  },
  {
    "start": 863.73,
    "end": 868.75,
    "text": "You can remove pedestrians, put in faster or more pedestrians, whatever you like.",
    "speaker_id": null
  },
  {
    "start": 869.27,
    "end": 870.35,
    "text": "And you have this data set.",
    "speaker_id": null
  },
  {
    "start": 870.43,
    "end": 874.11,
    "text": "You can feed into the AI training and be ready to deploy it.",
    "speaker_id": null
  },
  {
    "start": 875.15,
    "end": 877.85,
    "text": "Let's talk about how to shrink the domain gap.",
    "speaker_id": null
  },
  {
    "start": 878.449,
    "end": 886.709,
    "text": "The domain gap, but we also call it reality gap, is the gap between the simulation domain, the simulation world, the virtual world.",
    "speaker_id": null
  },
  {
    "start": 887.329,
    "end": 892.79,
    "text": "Our simulations take place and the real world where measurements take place on the street or in the air.",
    "speaker_id": null
  },
  {
    "start": 893.189,
    "end": 902.79,
    "text": "To bridge that domain gap in radar is extremely complicated because you'll need to make sure that you have all radar-specific aspects covered that do make an influence in the real world.",
    "speaker_id": null
  },
  {
    "start": 903.01,
    "end": 913.329,
    "text": "So hardware imperfections, non-linearity on other sensor effects, as well as world behavior like imperfect surfaces that radar experience in the real world and all these types of things.",
    "speaker_id": null
  },
  {
    "start": 913.43,
    "end": 935.329,
    "text": "And our USP is that we've built a processing chain where we can set up radar realistic scenarios, where we can build full digital radar twins that covered these aforementioned hardware effects and non-linearity imperfections and feed this into a radar simulator that was specifically designed to cover wave propagation effects that we'll need to cover for radar.",
    "speaker_id": null
  },
  {
    "start": 935.93,
    "end": 954.49,
    "text": "And this chain allows us to simulate radar raw data in a level of realism that is currently unavailable on the market and that we need to bridge the domain gap and to build the trust that virtual development really is an option to speed up development processes in real world applications.",
    "speaker_id": null
  },
  {
    "start": 954.53,
    "end": 954.949,
    "text": "So you can",
    "speaker_id": null
  },
  {
    "start": 955.55,
    "end": 955.85,
    "text": "basically",
    "speaker_id": null
  },
  {
    "start": 956.27,
    "end": 959.81,
    "text": "synthesize AI training data, expensive AI training data.",
    "speaker_id": null
  },
  {
    "start": 959.93,
    "end": 961.949,
    "text": "Can you give us a Rohde & Schwarz use case?",
    "speaker_id": null
  },
  {
    "start": 962.77,
    "end": 964.329,
    "text": "Yes, I can give you a use case.",
    "speaker_id": null
  },
  {
    "start": 964.37,
    "end": 969.39,
    "text": "And that's actually a use case that probably everybody who is hearing this podcast has experienced before.",
    "speaker_id": null
  },
  {
    "start": 969.49,
    "end": 972.01,
    "text": "And that is micro body scanners at the airport.",
    "speaker_id": null
  },
  {
    "start": 972.209,
    "end": 978.089,
    "text": "So, you know, these things where you stand, you spread your arms and the machine tells you, you're good to go, hopefully.",
    "speaker_id": null
  },
  {
    "start": 979.13,
    "end": 979.209,
    "text": "And",
    "speaker_id": null
  },
  {
    "start": 979.25,
    "end": 981.11,
    "text": "you're able to walk through now.",
    "speaker_id": null
  },
  {
    "start": 981.25,
    "end": 981.709,
    "text": "That's the next",
    "speaker_id": null
  },
  {
    "start": 981.75,
    "end": 981.829,
    "text": "thing.",
    "speaker_id": null
  },
  {
    "start": 982.17,
    "end": 982.39,
    "text": "Yes.",
    "speaker_id": null
  },
  {
    "start": 983.51,
    "end": 985.569,
    "text": "And what we have here is",
    "speaker_id": null
  },
  {
    "start": 986.17,
    "end": 994.11,
    "text": "also a radar sensor and probably the most complex radar sensor there is on the civil market a really complex radar.",
    "speaker_id": null
  },
  {
    "start": 994.23,
    "end": 1015.89,
    "text": "Then we have a really complex imaging algorithm that takes all of the radar data together to create a three-dimensional image from the body and then you have an AI that actually tells the system yes that's non-hazardous so there's no knives there is no problem on that human And then we can finish the scan, for example.",
    "speaker_id": null
  },
  {
    "start": 1016.67,
    "end": 1019.37,
    "text": "And that AI is extremely complex.",
    "speaker_id": null
  },
  {
    "start": 1019.65,
    "end": 1021.469,
    "text": "Yeah, it had been trained with a lot of people.",
    "speaker_id": null
  },
  {
    "start": 1021.81,
    "end": 1029.57,
    "text": "So just that I get it right, a benefit here would be when we train this security scanner at airports.",
    "speaker_id": null
  },
  {
    "start": 1030.79,
    "end": 1036.67,
    "text": "Nowadays, currently, we need real people walking through these scanners.",
    "speaker_id": null
  },
  {
    "start": 1037.35,
    "end": 1039.95,
    "text": "to feed AI with real data.",
    "speaker_id": null
  },
  {
    "start": 1040.39,
    "end": 1040.55,
    "text": "Real",
    "speaker_id": null
  },
  {
    "start": 1040.63,
    "end": 1042.45,
    "text": "people with real dangerous objects.",
    "speaker_id": null
  },
  {
    "start": 1043.31,
    "end": 1044.57,
    "text": "Yes, actually.",
    "speaker_id": null
  },
  {
    "start": 1045.33,
    "end": 1049.67,
    "text": "What you're offering is that you can simulate the whole situation.",
    "speaker_id": null
  },
  {
    "start": 1050.33,
    "end": 1073.89,
    "text": "And if you want to develop it further into the future, you can easily have virtual people walking through a virtual QPS security scanner with very virtual dangerous things on their bodies and still get data out of your virtual system for the real world, right?",
    "speaker_id": null
  },
  {
    "start": 1073.99,
    "end": 1074.77,
    "text": "Very good summary.",
    "speaker_id": null
  },
  {
    "start": 1075.03,
    "end": 1077.03,
    "text": "I think we've briefed you extremely well.",
    "speaker_id": null
  },
  {
    "start": 1078.07,
    "end": 1083.19,
    "text": "And just thinking about various use cases in the future, what pops up in my mind is space.",
    "speaker_id": null
  },
  {
    "start": 1083.83,
    "end": 1088.55,
    "text": "For this, obviously, you also need simulation because there you build super complex systems.",
    "speaker_id": null
  },
  {
    "start": 1088.59,
    "end": 1089.69,
    "text": "They are super expensive.",
    "speaker_id": null
  },
  {
    "start": 1090.43,
    "end": 1095.85,
    "text": "You bring them out to the outer solar system, they fly for several years and then they need to work.",
    "speaker_id": null
  },
  {
    "start": 1095.93,
    "end": 1098.37,
    "text": "You need to be sure that it is working.",
    "speaker_id": null
  },
  {
    "start": 1098.51,
    "end": 1114.23,
    "text": "So their simulation can be also really, really helpful to know before you build and then fly the system to some other planet that you know what you expect and therefore be prepared in your algorithms.",
    "speaker_id": null
  },
  {
    "start": 1115.29,
    "end": 1119.01,
    "text": "Now you are not like any kind of company sitting here with us today.",
    "speaker_id": null
  },
  {
    "start": 1119.99,
    "end": 1122.27,
    "text": "But you are a strategic partner.",
    "speaker_id": null
  },
  {
    "start": 1122.41,
    "end": 1127.55,
    "text": "You have a strategic partnership with our technology group, Rohde & Schwarz.",
    "speaker_id": null
  },
  {
    "start": 1127.91,
    "end": 1129.99,
    "text": "Actually, the pretty simple question is why.",
    "speaker_id": null
  },
  {
    "start": 1130.57,
    "end": 1137.75,
    "text": "Christina Gessner tells us, she is our executive vice president for the whole division test and measurement.",
    "speaker_id": null
  },
  {
    "start": 1138.23,
    "end": 1139.61,
    "text": "Inga has spoken to her.",
    "speaker_id": null
  },
  {
    "start": 1142.15,
    "end": 1144.91,
    "text": "The general topic of technology partnerships.",
    "speaker_id": null
  },
  {
    "start": 1145.27,
    "end": 1146.33,
    "text": "Why does it matter?",
    "speaker_id": null
  },
  {
    "start": 1146.65,
    "end": 1158.97,
    "text": "For us, it's really important to collaborate closely with universities, with research, throughout the ecosystem, to make sure we approach new technologies and their applications particularly.",
    "speaker_id": null
  },
  {
    "start": 1159.73,
    "end": 1168.53,
    "text": "And we have seen a lot of examples where this collaboration really resulted in excellent innovation, really successful innovation.",
    "speaker_id": null
  },
  {
    "start": 1168.79,
    "end": 1170.15,
    "text": "When we talk about Rohde & Schwarz.",
    "speaker_id": null
  },
  {
    "start": 1170.79,
    "end": 1175.43,
    "text": "We do have some kind of tradition of innovation for almost a century now.",
    "speaker_id": null
  },
  {
    "start": 1175.75,
    "end": 1179.71,
    "text": "What drives innovation when we work together with academia here?",
    "speaker_id": null
  },
  {
    "start": 1179.93,
    "end": 1180.15,
    "text": "It's to",
    "speaker_id": null
  },
  {
    "start": 1180.25,
    "end": 1182.21,
    "text": "make sure you bring those worlds together.",
    "speaker_id": null
  },
  {
    "start": 1182.39,
    "end": 1190.11,
    "text": "The ideas you have, either coming more from a product, a customer application point of view, then you have the technology view.",
    "speaker_id": null
  },
  {
    "start": 1190.91,
    "end": 1196.81,
    "text": "the new technology enables, respectively, and you bring this together for the benefit of a customer.",
    "speaker_id": null
  },
  {
    "start": 1197.07,
    "end": 1204.45,
    "text": "It's really about bringing ideas into real products or turning ideas into real products successfully.",
    "speaker_id": null
  },
  {
    "start": 1204.65,
    "end": 1206.51,
    "text": "A specific example would be fiveD.",
    "speaker_id": null
  },
  {
    "start": 1206.99,
    "end": 1209.73,
    "text": "We are covering in this podcast episode today.",
    "speaker_id": null
  },
  {
    "start": 1211.07,
    "end": 1214.37,
    "text": "Why is fiveD particularly important?",
    "speaker_id": null
  },
  {
    "start": 1214.53,
    "end": 1216.05,
    "text": "What makes it unique?",
    "speaker_id": null
  },
  {
    "start": 1216.83,
    "end": 1222.95,
    "text": "FiveD is really unique, reaching the gap between virtual world, simulation world and reality.",
    "speaker_id": null
  },
  {
    "start": 1223.19,
    "end": 1234.71,
    "text": "So when we talk about our development for security scanners, there's really large amounts of data required to do training of machine learning algorithms and so on.",
    "speaker_id": null
  },
  {
    "start": 1234.89,
    "end": 1249.39,
    "text": "So the approach that fiveD brings in here for their hyper realistic radar simulation is really addressing the topic and even widening it up for a lot of different use cases if you think of for example automotive radar.",
    "speaker_id": null
  },
  {
    "start": 1249.77,
    "end": 1262.93,
    "text": "The security scanners you mentioned is also a bit unique, because that is a product people around the world can kind of experience when they are at the airport to have their personal security check.",
    "speaker_id": null
  },
  {
    "start": 1263.33,
    "end": 1267.43,
    "text": "Usually, Rohde & Schwarz sits a bit behind the scenes.",
    "speaker_id": null
  },
  {
    "start": 1267.61,
    "end": 1271.13,
    "text": "We do test and measurement what you are a leader of.",
    "speaker_id": null
  },
  {
    "start": 1273.27,
    "end": 1282.19,
    "text": "How do other customers, partners, maybe benefit or have added value here through this partnership with",
    "speaker_id": null
  },
  {
    "start": 1282.27,
    "end": 1282.73,
    "text": "fiveD?",
    "speaker_id": null
  },
  {
    "start": 1283.15,
    "end": 1291.03,
    "text": "I think we are seeing this necessity to bridge this gap between simulation and real world in many different applications.",
    "speaker_id": null
  },
  {
    "start": 1291.17,
    "end": 1299.37,
    "text": "I mentioned automotive radar, but also if you think about wireless communications, there are similar challenges there.",
    "speaker_id": null
  },
  {
    "start": 1299.55,
    "end": 1307.69,
    "text": "So I think there's a lot of different ways and different possibilities how the solutions that fiveD is developing can help them.",
    "speaker_id": null
  },
  {
    "start": 1309.03,
    "end": 1314.59,
    "text": "Going forward for ten years, anything you may Imagine how this will evolve.",
    "speaker_id": null
  },
  {
    "start": 1315.45,
    "end": 1324.47,
    "text": "We would definitely like to continue this approach, continuing to emphasize the collaboration with academia, partnership with startups.",
    "speaker_id": null
  },
  {
    "start": 1326.27,
    "end": 1329.33,
    "text": "And there's a lot technology-wise coming up.",
    "speaker_id": null
  },
  {
    "start": 1329.45,
    "end": 1349.87,
    "text": "If you think about the next ten years, if you think about how AI evolves, if you think about quantum computing, for example, there's really a lot of technologies here that we are monitoring that have a huge impact on our business, on test and measurement business, and also provide a lot of opportunities for us to provide solutions.",
    "speaker_id": null
  },
  {
    "start": 1350.15,
    "end": 1354.09,
    "text": "So, therefore, from my point of view, it will just become more important.",
    "speaker_id": null
  },
  {
    "start": 1359.61,
    "end": 1369.43,
    "text": "When you look ahead and think of radar development, where do you see, let's say within the next years, the biggest changes and disruptions?",
    "speaker_id": null
  },
  {
    "start": 1370.53,
    "end": 1377.29,
    "text": "Especially driven by simulation, we can expect huge breakthroughs with respect to radar AI.",
    "speaker_id": null
  },
  {
    "start": 1378.25,
    "end": 1384.89,
    "text": "Because this is currently really limited on the availability of suitable datasets that are large, diverse and labeled.",
    "speaker_id": null
  },
  {
    "start": 1386.91,
    "end": 1388.65,
    "text": "Up to date still a lot of manual effort.",
    "speaker_id": null
  },
  {
    "start": 1389.47,
    "end": 1393.21,
    "text": "And I also want to add another point where also AI plays a big role.",
    "speaker_id": null
  },
  {
    "start": 1393.29,
    "end": 1399.49,
    "text": "And this is the current trend in several industries, for example, automotive or aerospace.",
    "speaker_id": null
  },
  {
    "start": 1399.53,
    "end": 1404.19,
    "text": "Defense is a software defined scheme that's coming more and more where the approach is.",
    "speaker_id": null
  },
  {
    "start": 1404.25,
    "end": 1405.19,
    "text": "you have a hardware.",
    "speaker_id": null
  },
  {
    "start": 1405.51,
    "end": 1410.29,
    "text": "The hardware is updated, not yearly, but several years basis.",
    "speaker_id": null
  },
  {
    "start": 1410.67,
    "end": 1414.13,
    "text": "But the actual software is changed regularly.",
    "speaker_id": null
  },
  {
    "start": 1414.57,
    "end": 1415.99,
    "text": "It's evolving",
    "speaker_id": null
  },
  {
    "start": 1416.79,
    "end": 1417.53,
    "text": "much quicker.",
    "speaker_id": null
  },
  {
    "start": 1418.11,
    "end": 1422.27,
    "text": "And for this, you need to be much faster in development and in the testing process.",
    "speaker_id": null
  },
  {
    "start": 1422.67,
    "end": 1426.21,
    "text": "And here again, simulation can play a really vital role.",
    "speaker_id": null
  },
  {
    "start": 1427.41,
    "end": 1429.89,
    "text": "So thank you for being here today.",
    "speaker_id": null
  },
  {
    "start": 1429.97,
    "end": 1432.89,
    "text": "When I listen to you, I see that you're excited.",
    "speaker_id": null
  },
  {
    "start": 1432.99,
    "end": 1434.65,
    "text": "You're deep into the tech.",
    "speaker_id": null
  },
  {
    "start": 1435.95,
    "end": 1444.05,
    "text": "Actually, that is kind of how I imagine the founders of Rohde & Schwarz meeting at the university level.",
    "speaker_id": null
  },
  {
    "start": 1444.71,
    "end": 1449.55,
    "text": "Doing research together, thinking about customer problems to solve.",
    "speaker_id": null
  },
  {
    "start": 1450.59,
    "end": 1461.19,
    "text": "It feels a bit like this start-up spirit still lives within Rohde & Schwarz, although it's a pretty huge technology group worldwide nowadays.",
    "speaker_id": null
  },
  {
    "start": 1462.39,
    "end": 1477.53,
    "text": "And these corporations, with for example, fiveD and Michel Imacelle, who joined the show today, is for me the best kind of proof that innovation still drives the group.",
    "speaker_id": null
  },
  {
    "start": 1478.21,
    "end": 1478.97,
    "text": "Thank you for sharing.",
    "speaker_id": null
  },
  {
    "start": 1479.05,
    "end": 1479.87,
    "text": "Thank you for being here.",
    "speaker_id": null
  },
  {
    "start": 1480.55,
    "end": 1481.21,
    "text": "Thanks for having us.",
    "speaker_id": null
  },
  {
    "start": 1481.67,
    "end": 1482.83,
    "text": "Thank you very much for your invitation.",
    "speaker_id": null
  },
  {
    "start": 1482.89,
    "end": 1483.43,
    "text": "It was great fun.",
    "speaker_id": null
  },
  {
    "start": 1485.43,
    "end": 1493.81,
    "text": "And to all our listeners, thank you for tuning in and don't forget to subscribe to the podcast and to look in the show notes,",
    "speaker_id": null
  },
  {
    "start": 1493.93,
    "end": 1498.91,
    "text": "there you will find some additional material and contact information.",
    "speaker_id": null
  },
  {
    "start": 1500.19,
    "end": 1503.79,
    "text": "So hear us next time and don't forget to subscribe.",
    "speaker_id": null
  }
]