For his Hackaday Prize entry, [having11] is building a simple and cheap fall detection notification button that can be worn by senior citizens, young kids, or others affected by medical conditions. [having11] did some fact-finding, and it appears that falls are one of the leading causes of injury for seniors, according to data available from the Centres for Disease Control and Prevention (CDC).
This device will sense a fall and send a text message or email to a recipient caregiver, loved one, or friend. The notification can also be manually triggered using a pushbutton. There’s a 5-second delay before it actually sends the alert, allowing a false trigger to be canceled. On receiving the alert, the recipient can decide how to proceed and if the situation requires a call to emergency services.
The device uses an ESP8266, a MPU6050 MEMS gyro-accelerometer combo, and MyDevices Cayenne IoT service. The Cayenne IoT service is free for Makers and non-commercial use at the moment. The only other components needed are a few discretes and a small LiPo battery, keeping the cost of the device under $10. The whole assembly is housed in a 3D-printed enclosure. The next steps would probably be to make it more compact and design a housing that can be worn as an arm or chest band or on a waist belt. An important requirement of such monitoring devices is being able to notify when/before it is unable to fulfill its primary requirements. Towards that end, maybe adding a low battery and low WiFi signal strength indicators would be nice.
If you have more suggestions on making this better, chime in with your comments below.
That could be applied to man down duress alarms current ones used in the facility I work in onlyboperate in one direction
Theres a good reason for it only operating in one direction, and that is to prevent nuisance alarms. Remember that for a duress alarm, you cannot have a “press the button to cancel the alarm” thingy, because then the bad guys could cancel it after having beaten down the person in question.
Many duress alarms are designed to be placed in the belt, and that can cause problems with sitting down in a chair for example. Thats why duress alarms must rather have false negatives than false positives, eg better to not to send a alarm in a bad situation, than to send a nuisance alarm in a innocent situation (and cause a unneccesary SWAT summon). After all, a duress alarm that sometimes fail to transmit a alarm in a real situation, but mostly works, is better than nothing.
Good effort, but make it wireless rechargeable to that they can take it off and swap it over for a second one that is sitting on the charger and ready to go. Also may the charge state very obvious even to a person with hearing and vision loss.
If you have a plan for making any money from your device, be sure to follow up on what is going on between Nintendo and iLife. https://arstechnica.com/tech-policy/2017/09/jury-finds-nintendo-wii-infringes-dallas-inventors-patent-awards-10m/
Nowadays even cheap smartwatches host some connectivity and an accellerometer.
I really like the enclosure and effort.
Similar boxes can also be used to monitor goods if are shaken too much, measure throwing performance, burglar alarm, drone monitoring and black box, measure inclination and stress of bridges, emergency sysmographers, step counters and much more.
5 seconds? Person probably will have a fit trying to scramble for the cancel button.
Increase it to 40 seconds at least, it’s not like it matters that much.
Hopefully they only fall near their wireless router?
Yes. Problem ensue when elderly people fall at home without say a partner around to call for help. If they fall in the supermarket, you should think someone may volunteer to help them up or call emergency services if required….
Surely they can just call 0118999881999119725…3?
Newer ambulances, faster response times, and better-looking drivers!
I think a accelerometer/gyro is the wrong type of sensor to use as fall detection. Its better with a tilt sensor instead, and have it mounted on the body. (Perhaps with a extension lead to the arm-mounted button and then a velcro strap around the chest with the tilt sensor mounted in such a way it wil detect the body being in horizontal or near-horizontal direction)
Combine that with a bed sensor (A mat you put in the bed and detects if a person is in the bed) that inhibits alarm when triggered, and you got a perfect solution. (Eg, when the button detects tilt – check if bed sensor is ON, then don’t send alarm)
There are tilt sensors available that has a cone and a ball bearing, and as soon as the ball bearing touches the roof above the cone, it will trigger, which will gurantee detection in all 360* fall angles. Think a less sensitive variant of the sensor used for towaway detection in car alarms. The one used for car alarms are wayy to sensitive however.
I think the current accelerometer solution will nuisance trip very often, causing the wearer to stop wearing it, like the original problem. With “nuisance trips”, I mean that it wil trigger way too often so the old person have to “cancel” his alarm within the 5 second window all the time.
Maybe the sensor works for some people. It depends on how they use their body and arms. But I imagine that a person that even does small things like drive a nail into a wall with a hammer or just chopping some food with a knife, would trip it.
The problem with elder people and emergency buttons:
– They forget to wear it
– They loose it somewhere
– They forget to change the batteries
– In case of an emergency, they forget that they have it
– They keep wearing it in the shower, or it ends up in the washing machine
– etc.
My Sister In Law lives with us. She had a fall on fri. and it was 5 min before we saw her on the floor.
I had been thinking of something like this for a bit now. Not so much for my family but for me.
So I guess I better get moving on making some of these.
Thank you for your work, I have not read all your work yet. I will be looking into what you have done and see if it is the same as what I’ve been thinking.
Good job Thanks.
Excellent idea. However, CDC & P (The P for prevention is implied but not on the sign in Atlanta.) spells its name using the American spelling for CDC, that is Centers, not the English one.
Yeah, and every false alarm this thing sends triggers an emergency service response – which costs thousands of dollars per visit – whether you need it or not. The cost is NOT covered by insurance of-course. So the user simply turns the damn thing off to avoid going broke. Epic FAIL! All this assumes the user is a Taxpayer of-course. More on that below…
There was a day in the U.S. where almost all citizens would pay TAXES and police and fire would come when needed, without BILLING anyone. No more. Now roughly 50% of everyone in the U.S. actually pays TAXES, and the remaining 50% are Freeloaders.
So, if you call the police or fire department today, expect to get a BILL. Of-course, the people who DON’T pay taxes do not have to pay the bill, but the people who DO pay taxes do have to pay the bill. So the taxpayers subsidize the non taxpayers, as it should be – but NOT at a 50% taxpayer to 50% non-taxpayer ratio! What’s gong on here is termed “Wealth Distribution”, and Comrade, you know that when you redistribute wealth to those that didn’t earn it – you Buy Their Vote.
Hi, it is having11. This device simply alerts a CAREGIVER or other person you want, so they can then call 911 or come and check on you. This saves money. It’s a one-time cost, there’s no monthly fee, and the fire department doesn’t show up every time it detects something. I agree with your stance on freeloading, I dislike people who just take advantage of the system. I don’t think that has to do with my project, but OK.