Hackaday Prize Entry: Vibhear

Hearing impairment, either partial or total, is a serious problem afflicting a large number of people. Almost 5% of the global population has some form of hearing disorder. For those affected by this disability from birth, it further impacts the development of language and speech abilities. In recent years, cochlear implants are increasingly being used to address this problem. These implants consist of two parts – the receiver and electrode array are implanted under the skin near the ear (with the electrode array terminating inside the Cochlea), while the microphone, electronics, transmitter and power source are attached on the outside. Often, the external unit has to be removed – for example, when the person needs to sleep. This is particularly so in the case of young children. The external unit is fairly large compared to their head and causes discomfort during sleep. And parents are worried that the expensive device could get damaged when the child is sleeping. This leads to the alarming situation where the child is asleep and has no audio sensory inputs being received from the surroundings. Not only can they not hear morning alarms, but also cannot react when there is an emergency situation such as a smoke alarm going off.

[Srdjan Pavlovic] came across this problem first hand when he visited his friend and learned about their six-year-old son with hearing loss since birth. The parents said their child will not be disturbed by loud noises at night since the external unit of his cochlear implant is removed each night. [Srdjan] then started work on building the Vibhear – an assistive hearing device to be used when the main hearing aid is removed or not working. It is a low-cost arm-band that provides a vibratory signal in response to high ambient noises.

The main components are a microphone, amplifier, microcontroller and vibration motor powered by a LiPo battery through a boost converter/charger. An RTC module allows setting up daily wake up alarms. It’s currently prototyped around the Arduino, but the next iteration will use a specialized DSP which can be programmed to perform signal processing operations on input sound. This will allow identification of specific sounds such as car horns, barking dogs, smoke alarms or emergency sirens.

[Srdjan] is in the process of choosing components for his next iteration, so if you have any recommendations to help him choose the microcontroller, power supply controller or other parts, do let him know via comments below.

18 thoughts on “Hackaday Prize Entry: Vibhear

  1. A couple of things to consider: Would the device still work correctly if the microphone were to be covered by sheets/pillows or even the patients own body? Any false positives from things rubbing against the microphone? Commercial variants have the microphone more off to the side on a nightstand with the indicator underneath the pillow.

    Otherwise, I like this implementation: https://www.cnet.com/news/wasabi-smoke-alarm-raises-a-stink-in-japan/

  2. Born deaf here, my hearing loss was 95dB at low frequency to 125dB at 3KHz so I couldn’t hear high pitched alarm or those using piezo element Cochlear didn’t become widely available until I was already in my 20s. Not that it mattered, I’ve slept with regular hearing aids and slept through loud noises so sound still won’t work on heavy sleeper.

  3. Hi Sam,
    You hit right into the center!
    We came upon same question while discussing effectiveness of microphone so we plan to test this use case thoroughly.
    If you are interested in test results, we will publish all results within project logs.
    Thank you for your comment.

    1. I’ll keep an eye on your project. It will be interesting to see it come to fruition.

      One thing I want to suggest is to have the microphone a separate unit so that you won’t have to deal with the problem of the microphone being covered up. You can have the sensor unit be part of a plug-in night light which will then send a signal to your wearable to vibrate/do something else upon detection. That way, you will only need to deal with a single battery powered device. Energy consumption by the wearable can be minimized as it only has to act as a receiver and be sleeping most of the time.

  4. Human audible range to low frequency Haptic transducers are nothing new. In-fact some of the first designs simply convert normal human audible sounds to subsonic and output to a low frequency transducer like a subwoofer. Nothing to wear. No batteries to recharge.

    In a room, this eliminates the need for anything the hearing impaired person needs to wear. A decent HiFi set with speakers can be used. A smoke alarm beeping is converted to a HiFi speaker thumping. Same goes for a dog barking or even someone knocking on the door.

    There is lots that can be done with this NO-WEARABLE approach to audible to sub-audible haptic conversion, especially if modern wireless DSP enabled embedded processors and beam-forming array microphones are employed.

    Gee, in the last paragraph above didn’t I just describe the likes of one of these Google or Amazon “Spy-Speakers”, like “Google Home” that are so popular today? Yep – I did. so now expect tons of frivolous patents to crop-up where these horrible spy devices warn the hearing impaired based via subaudible signals.

    For example what the hearing-impaired person “hears” or better-yet “feels” is a down converted version of the ambient sound. For-example, a smoke alarm (sold by Google of course) or a doorbell (Sold by Google of-course), or even a barking dog (currently NOT sold by Google, but it’s only a question of time). etc., etc. By using sub-sonic feedback either by making low frequency sound through an existing audio device, and/or by sending low frequency sound via wireless to your home multimedia device and/or connected TV (or toaster, or refrigerator, or toilet, or washing machine, all controlled by Google of-course…)

    Of course there’s nothing stopping the haptic signalling from going to a “wearable” device. For example a “Google Wear” smart watch! (“Smart” watches are still IMO an EPIC FAIL!)

    Yeah, the bad idea in all this is the likes of Google – who IS EVIL. Everything heard will be heard by Google too!

    Say you heard it here first….

  5. Hi Genki,
    Thank you for first-hand information.
    While testing if bracelet vibration is going to wake up its user successfully, we have noticed that heavy sleepers need stronger vibration.
    Do you have experience with some similar kind of assistive hearing device?
    What is your opinion about using vibration to draw person`s attention?

    1. My alarm clock uses strobe light to wake me. Vibration doesn’t work well as I sleep with a few cats and they often play on my bed. Clock are plentiful, party strobe light are also easy to get cheap. It’s a matter of wiring a relay from alarm clock to 110v mains to power the strobe light. Clocks with built in power plug used to be common in 80s and 90s but with everything like coffee machine, TV, radio, etc all having their own clock and automatic wake or start brew features, most clocks maker stopped offering this feature so it’s either the tiny unit with round dial and no snooze or hack myself one.

      Smoke alarm also has strobe light to wake me. It is connected to mains and has battery backup in case of power failure. I used to use teletypewriter for the deaf to (commonly referred as TDD) communicate but it has been about 15 years since I used one. Payphone were disappearing which meant if I had a car problem or needed something, I may have to walk miles to find a phone. I switched to using cell phone which has SMS and since almost everyone has one, I don’t need a relay service to connect to people who doesn’t have a TDD.

      I do not have special doorbell, it’s plain old chime. If I am up, I’m wearing hearing aids anyway and can answer door if needed. If I’m not wearing hearing aids, then I am probably not interested in whoever’s behind the door or is sleeping.

      The biggest challenge I’ve faced so far in the last few years are the online games. Many of the MMORPG are using voice chat system for strategies and control. Raid leader knows I am deaf and they have to type out strategy guide and instructions if I need to do something specific in an encounter, then hope I can deal with the unexpected when it’d be hard to take 15 seconds to type out something like “tank’s down Genki move in” or “rez the dead guy”

  6. I had a few thoughts:
    1. instead of usual haptic alarm, use a mild shock, which easy to generate (tear apart one of those cheap electric exercise stimulators you can get for less than $10 from China).
    2. Use an external speaker with a BLE interface to the person’s receiver. Then there’s no worry about muffling the mic. Also, with BLE, other systems can be readily coupled like fire/smoke alarms, laundry, door bell, etc.
    3. I read some time back of energy harvesting from external sources, especially WiFi. That could be used with a small supercap to power a BLE enabled micro like the ones used in wrist band devices.

  7. I use an alerting device -Bellman & Symphon. This can be linked to the doorbell, smoke alarms, telephone and an alarm clock. -The latter I do not use as I use a lamp on a time switch to wake up in the morning. The B & S pager can be worn around the neck or clipped to a pocket. It vibrates when activated. I sleep with it under my pillow and find it most effective. This is worth considering as an alternative to your system.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.