sonoware GmbH

We are excited to be part of the network! 👏

As the year slowly but steadily comes to an end, we want to take the time to express our gratitude to all our clients, partners, friends and families for their trust and continuous support throughout the year. We wish you all Merry Christmas, happy holidays and a great start into 2025. 🎅 🌲 See you next year! Your sonoware GmbH team!

What’s the status inside our bones? A quick dive into our acoustical multi-path bone investigator, or in short: AMBIT 🦴 Last Friday, our colleague Finn Spitz presented the “first fruits” of our collaborative project AMBIT at VDE DGBMT “Nacht der Biosignale”. Let’s dive into the promising development results of our osteoporosis detection device using nothing but ultrasound and acoustic signal engineering. Our bones consist of two elements, namely trabecular structures and the cortical bone. Latter are responsible for the strength and stability of the bones and are directly affected by osteoporosis. Here, the disease creates an imbalance in bone formation and breakdown, gradually increasing the bone porosity. To detect the density of the bone, the AMBIT device in development emits quantitative ultrasound signals at a critical angle below the bone surface, which are then measured and evaluated by a receiver. Here, the speed of sound, frequency and attenuation, for example, are then evaluated for diagnostic indications. To learn more about the functionality of our osteoporosis detection device, check out the recorded live stream (only in German) by VDE DGBMT: https://lnkd.in/effWRc2Q The project is conducted by sonoware GmbH and Kiel University Molecular Imaging North Competence Center (MOINCC). #osteoporosis #innovation #ultrasound #signalprocessing

Thank you for having us! It was a great pleasure to talk with you about our health-tech applications and we are very excited to see the development of our ultra sound signal processing software unfold their full capabilities. 🤩

Want to learn more about how to estimate the density of bones using nothing but ultrasound? 🦴 Dive into the podcast episode of VDE DGBMT to listen to our corporate story as well as the lessons learned from our research project with Kiel University. 🎙️

So this happened! 🎙️ Today, we were invited by VDE DGBMT to share a few insights into the measurement of bone density using ultrasound and extensive signal processing to detect osteoporosis. On November 15, our colleague Finn Spitz will be talking about the measurement of bone density beyond radiology at VDE conference "Nacht der Biosignale". Want to learn more about what we have been doing in health tech? Check out the podcast or join us on November 15 at Kiel University Technical Faculty. Conference: https://lnkd.in/dq8bVkeA The episode will be published in the upcoming days at: https://lnkd.in/g64D7UP7 #osteoporosis #podcast #VDEDGBMT #sonoware Christian Lüke Finn Spitz Reinhard Barkmann

How do you detect almost invisible objects in real-time? Yesterday, our colleague Steffen Yader Stadler presented our lessons learned after the first three months of the project GhostNetBusters. Here are some insights: 1. Differentiating between seafloor and objects at the edge of the water column seems to be difficult for most bottom line detectors. While the human operator is aware that the bottom hasn’t risen several metres, machine learning algorithms need to be taught to recognize this sudden rise as an object of interest. 2. Objects in the gain range are “just” invisible to the human eye. As the range of the sonar signal increases, the increasing gain makes human annotation a serious problem because objects are hard to see on the image. However, since we know where the object should be, through geo-referencing and active data generation with our partners, we can train our algorithms to see beyond the acoustic noise and detect the invisible. 3. A side-scan map without water columns helps to distinguish between known and new objects. Wrecks, large boulders or seaweed do not change position underwater. Lost fishing gear, on the other hand, will move, wraps itself around these structures or crawls into the seabed. By correcting for slant range, we are able to create a holistic map of the seafloor. Nothing extraordinary, but quite useful. 4. At a certain distance, the beams overlap and cover the same area twice. While the quality of the signal decreases dramatically with distance, the spread of the beam results in multiple coverage of the area of interest. Through inverse modelling, we expect to gather more information about objects hidden in the noise, allowing them to be detected despite their invisibility. We look forward to sharing more exciting insights into the progress of the project in the coming months. The project is carried out by GEOMAR Helmholtz-Zentrum für Ozeanforschung Kiel and sonoware GmbH, funded by the AI fund of the Land Schleswig-Holstein. Maritimes Cluster Norddeutschland e. V. Wirtschafts- und Wissenschaftspark mariCUBE Warner Brueckmann Mia Schumacher

Today we are at the 5th AI Conference of the Land Schleswig-Holstein. Learn more about how to detect objects underwater with machine learning. Make things visible, which are hidden in poor quality sonar data and acoustic noise. Meet GEOMAR Helmholtz-Zentrum für Ozeanforschung Kiel and sonoware GmbH at booth A15 at Kiel University Audimax. The project GhostNetBusters is a collaborative project by Geomar and sonoware, funded by Land Schleswig-Holstein. Associated partner is One Earth - One Ocean e.V. who have recovered provided the lost fishing nets displayed at our booth.

Looking for Ghost Nets in the Kiel Fjord ⚓ Few weeks ago, sonoware GmbH along with the partners from GhostNetBusters, navigated the Kiel Fjord to scout the coastlines for lost fishing gear and ghost nets polluting the waters around the beaches and harbors. Thanks to the “SeeKuh,” owned and operated One Earth - One Ocean e.V., we were able to maneuver close to the shore and mark numerous points of interest. These locations will now be inspected by volunteering divers, who will confirm and remove the detected ghost nets. To detect these ghost nets, we used two “Starfish” side-scan sonars which were provided by Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research and MacArtney Underwater Technology Group. The materials in the ghost nets appear differently in the sonar images. Metal parts, such as lead lines and ground weights, show up as stones and glowing lines within the images. The plastics and fabrics of the nets themselves are harder to detect, as these materials have very limited sound-reflecting properties and are barely visible through acoustic detection. This is where sonoware’s signal processing technology comes into play. We look forward to sharing more details about the project's progress in the coming months. The project team consists of the GEOMAR Helmholtz-Zentrum für Ozeanforschung Kiel Helmholtz Centre for Ocean Research Kiel and sonoware GmbH, and it is funded by the AI fund of the state of Schleswig-Holstein. #oceanresearch #sonoware #digitalsignalprocessing Picture Credit: sonoware GmbH

The Challenges of In-Car Communication: Audio Signal Processing in Modern Vehicles (2/2) 🚗 In the previous post, we’ve introduced the various challenges in-car communication (ICC) systems have to tackle to enable seamless interaction between passengers and driver. Each of the challenges described requires unique hardware and software solutions and adapted algorithms to create the perfect conversation experience. Let us dive in some of the features, we are working with: 1.   Noise Reduction and Echo Cancellation One of the primary functions of audio processing in ICC is noise reduction. By using advanced algorithms, the system identifies and suppresses background noises, such as road noise or wind, ensuring that conversations remain clear and intelligible. Echo cancellation is another critical feature. It prevents the annoying echo effect that can occur during conversations, ensuring a smooth and natural interaction between passengers. Echo cancellation is particularly challenging, as it requires significantly lower latency and needs to distinguish between words spoken by persons and words repeated through the audio system. 2.   Automatic Gain Control ICC applications use something called automatic gain control (AGC) to balance the audio levels of different speakers. Whether someone is speaking softly or loudly, AGC adjusts the volume to maintain consistent sound levels, ensuring everyone is heard clearly. 3.   Directional Microphones and Beamforming Directional microphones and beamforming technologies focus on the sound from specific directions, typically where the speaker is located. This selective hearing capability minimizes the interference from other sounds and improves the clarity of communication within the vehicle. 4.   Adaptive Algorithms for Real-Time Adjustment Modern ICC systems use adaptive algorithms that continually adjust to changing acoustic environments. Whether you’re driving through a quiet suburb or a bustling city, the ICC system dynamically optimizes the audio for clear communication. If you ever have the chance to drive a traditional classic car, take a moment to enjoy the difference between the sound systems of the past and what your current car is able to do in regards to communication. The more modern your personal ride, the more likely you have gotten used to modern ICC systems. You might even have enjoyed our software without your knowledge. #sonoware #InCarCommunication #AudioSignalProcessing Fotocredit: Foto von ThisisEngineering auf Unsplash