Open Your Ears

If you are the type of person that always wants to have the latest gadgets, and as a Hackster News reader you likely are, then you may find yourself a bit disappointed with the latest commercial offerings. Sure, smartphone cameras get better every year, and smartwatches are getting incrementally better as well, but where are the big leaps forward? Where are the new platforms that transform important aspects of our lives?

Many people think that the next frontier will be with devices called earables. These are devices that go in, or near, the ears and provide advanced functionalities beyond basic audio inputs and outputs. But while earables have been an active area of research, that work has not materialized into many commercial products. That leaves this area wide open for innovation.

A group led by engineers at the Karlsruhe Institut of Technology in Germany see a huge opportunity for the development of transformational technologies in earables, but believe that the field is presently being held back by a lack of access to adequate development platforms. And the platforms that do exist tend to lack the extensibility necessary for real innovation. In response to this state of affairs, they have developed an inexpensive and highly extensible development platform called OpenEarable.

The initial version of the OpenEarables system was recently unveiled, and consists of a wearable platform instrumented with a three-axis accelerometer and gyroscope, an ear canal pressure and temperature sensor, and an inward facing ultrasonic microphone and speaker. To process this sensor data and handle interactions with the wearer, an implementation of the Arduino Nano 33 BLE Sense microcontroller development board was created on a custom over-the-ear hook-shaped PCB.

Leveraging the Arduino platform makes OpenEarables easy to work with via the Arduino IDE, and also allows developers to leverage existing libraries and documentation. This choice also keeps costs very low — about $40 per unit — which increases accessibility of the platform.

Firmware was developed for OpenEarable that captures data from all of its sensors and transmits it wirelessly via Bluetooth Low Energy. Support for machine learning on the edge was also built into the firmware — it includes functionality to collect and label data, and also to train and deploy machine learning models. This software was developed using the Arduino development ecosystem, which makes it simple for others to modify for their own purposes.

While OpenEarables was designed to be a development toolkit, and not a finished product of its own, the team still wanted to show what it is capable of, so they built some applications to demonstrate the system in action. In one example, they showed how the ear canal pressure sensor can be utilized to detect jaw movements, which has applications in recognizing chewing events, or even speech. Another demonstration showed how the ultrasonic microphone can pick up inaudible signals from the speaker, which can reveal the ​​shape of the ear canal. This signature can be used as an authentication mechanism for various services.

But of course these are just a few examples. What is really possible with OpenEarables is up to the developers that put this system to work. And the possibilities can be expected to grow in the future as the team is working on a number of enhancements at present. They want to upgrade some of the sensors, like the accelerometer, and also make it possible for the earpiece to be paired with another one on the other ear.

The OpenEarable hardware design and software is open source and available on the project website under a permissive license.