Hello everyone, and happy August!
As you already know by now, I’m working hard to look for interesting insights and products from the industry, and this edition is dedicated to acoustic detection.
The range of potential applications of acoustic detection is pretty wide, and we can find early examples that date back to the 1930s in the UK and predated radar. Back then, “sound mirrors” were placed on the British coast so observers could hear incoming enemy bombers, providing detection at night and in bad weather. You can see what they look like in the image below, or watch this YouTube video. See more bizarre pre-radar era acoustic detection devices from Europe.
In this edition, we’ll look into three interesting applications of acoustic detection and some future implications. It feels a bit like a technology from history is coming back, but in a very different shape and form.
Happy Reading! 📖
If you’re interested in the future of information, intelligence, and decision systems, I invite you to stay connected and subscribe to The Detectionist now.
📶 Interesting Signals to Watch
🚀Application 1 - Detecting hypersonic missiles
The application: In recent years, China, Russia, and the US have begun developing hypersonic missiles that fly inside Earth’s atmosphere, below the coverage of long-range radar. Unlike ballistic missiles, their trajectory is unpredictable, and that makes them harder to spot using radars.
HyperKelp, a California-based company, is developing special acoustic detection devices that are mounted on buoys inside the ocean. Their technology is already capable of detecting sonic booms, offering an alternative measure to detect hypersonic missiles. In this article, Dr. Graeme Rae, HyperKelp’s founder and CEO, explains more about the application of the technology and how warfare could change as a result.
🌿 Application 2 - Monitoring plant health
The application: Did you know that you can “listen” to plants?
Sonicflora is a Swedish startup developing technology that “listens” to the ultrasonic sounds plants emit to determine stressors related to their health.
The company is in the process of collecting data about ultrasonic sounds that plants make to determine if they suffer from lack of sun, water, viruses, etc., all before they show visible signs. Nowadays, they are focusing on helping tomato and cucumber growers detect problems within greenhouses, and they are working on potentially expanding this technology to more use cases in the future.
💦 Application 3 - Water leak detection
The application: Acoustic water leak detection is a non-invasive method that uses specialized microphones and listening devices to listen for the specific sounds that are created when water escapes from pressurized pipes. You can then see these specific sound waves on a screen or headset. Watch this video to see how this is done.
Companies such as EnviroTrace (Alberta, Canada) demonstrate the potential of this application at the municipal level, across large water systems.
🔮 What this could mean for the future
Why now: There’s a clear underlying driver here: the combination of progress in AI/ML models around sound identification and lower-cost listening devices makes acoustic detection a viable solution to detect blind spots across industries.
Implications: This is an interesting trend around applications of acoustic detection, and this is why I’m not analyzing each signal independently, but as one consolidated trend.
The common thread isn't the use of "sound" for detection. It's that acoustic sensing now detects things beyond what other technologies such as radars can detect. We can assume that more use cases will become visible in industries that have blind spots and where the stakes are high, such as defense, utilities, oil & gas, and more.
Potential future developments and opportunities:
For startups, this is the area worth looking into: wherever a domain relies on a dominant sensing method with known blind spots, acoustic detection may now be a viable, non-invasive, low-cost complement. The range of applications is yet to be discovered.
The real moat is proprietary datasets that detect a specific problem and that pose a real opportunity for companies with the skills to create such datasets as well as usable solutions around them. Cross-modal applications could emerge in certain industries (e.g., audio + visual sensing).
In addition, the dual-use nature of such applications could lead to export control and other restrictions due to the changing dynamics around geopolitics.
🙏🏻 A quick ask from you - Product and Product Marketing Leaders
I’d love to chat with those of you in charge of product or GTM roadmaps at startups anywhere from seed to Series B-C.
Let me know if you’d like to take 15-20 minutes online to chat about some of your workflows. I’m developing a workshop/course around these workflows and want to know how to better serve product and GTM leaders.
Feel free to reach out to me on LinkedIn or through info@yaroncohen.com
🔐 Protect your data and online reputation with Incogni.
If you Google your name, you’ll see how easily the world can find information about you. Hidden layers can be created by sketchy data brokers, exposing you to online threats. Incogni automates the process of removing your data from places it does not need to be found.
If you’re building something in the field of information, intelligence, and decision systems anywhere in the world, I want to hear about it. Email me at info@yaroncohen.com or connect with me on LinkedIn.
If you’d like to support my work and buy me a coffee, I invite you to do so with the button below.
If you know someone who works on solutions in this field, feel free to share this article with them.
Thank you very much, Yaron, for sharing such an interesting article.
It's fascinating to see acoustic sensor networks playing an important role again. They are now being widely used in Ukraine to detect incoming Shahed drones.
This technology has surprisingly deep roots. During World War II, before radar became widespread, the Soviet Union extensively used acoustic aircraft detectors to detect approaching German aircraft. Skilled operators could often estimate an aircraft's direction, engine type, approximate altitude, and even whether they were hearing a single aircraft or an entire formation simply from its sound signature.
The concept was simple: large acoustic horns collected engine noise, operators listened through headphones, determined the aircraft's direction, and relayed the information to searchlight and anti-aircraft units. It required considerable training, and weather conditions such as wind, rain, or temperature inversions could significantly affect performance.