When you travel, you likely collect photographs and knick-knacks that can be displayed nicely for yourself and others. David Levin, however, took this one step further and used an MP3 recorder to capture the sounds of the places he and his wife have visited. But how does one show off sounds? Levin has a clever answer for that in the form of his Arduino-based Audio Memory Chest.
The project uses a recycled card catalog to hold items from each place traveled, and when one drawer is pulled out, a magnet and Hall effect sensor tells an Arduino Pro Mini which drawer has been opened. A serial MP3 player module then produces a random audio file recorded at that location, treating the user to both the sights and sounds of the region!
My wife and I have been lucky enough to travel all over the world together during the past few years. Wherever we go, I collect little knick knacks, souvenirs, and ephemera. I also use a little MP3 recorder to capture sounds (marketplaces, street sound, music, etc). It’s always amazing to listen to these later—they immediately bring you back to a place, far better than a photograph alone could.
On this week’s Internet of Things Podcast (play or download above), Stacey Higginbotham and Kevin Tofel give a hands on review of the Amazon Echo’s new multi-room audio feature and provide some tips for you to get started using it.
They also discuss how GE is scaling back its industrial IoT dreams, explain what a partnership between Amazon and Microsoft’s Cortana means, and the Elgato sensors.
There’s a new Internet of Things Podcast Listener Hotline, and Kevin explains how he built the system using a Raspberry Pi!
This week’s guest is Alasdair Allan, a tinkerer and researcher who is thinking about the way we secure highly distributed systems. Stacey and Alasdair discuss how malicious data inserted into a system that can report false information to bring about a destructive action.
Audio Analytic, the Cambridge-based Artificial Intelligence company operates an intelligent sound recognition software it calls ai3. Based on advanced machine learning techniques, the software recognizes significant sounds in smart home devices and takes automated action.
Founded at Cambridge, UK in 2008, the company has its U.S office in Palo Alto, CA. It has secured total equity funding of $ 8.12M in 3 Rounds from 5 Investors. The latest round was $ 5.5M Series A in Jan, 2017 led by Cambridge Innovation Capital.
Another feature of the software is audio anomaly detection which is understanding the normative pattern of sounds within an individual home and sending an alert whenever deviations occur. The sounds can be like the clatter of something falling, or the hiss of water pipe.
The software works on embedded systems such as ARM, X86 and MIPS. It is compatible with operating systems Linux®, Windows®, Android® and RTOS. The company recently partnered with Angee, an autonomous home security system to bring audio artificial intelligence to the security system.
A major differentiation of the AI-based sound recognition software is that it is creating the taxonomy of all sounds which makes it sound recognition module much better than Apple Siri, Google Home and Amazon Alexa. The company is working on addressing the ‘false alarm’ challenge. It holds a patent in reducing false positive identification of sound.