Pervasive Health Monitor (Got Granny?)

[Reza] sent in a project that he’s obviously put loads of work into. His Pervasive Health Monitor is basically a bluetooth enabled health telemetry recorder/transmitter. I think it’s an absolutely excellent piece of work. He’s offered to post more technical details if we have enough interest – It’s got my vote.

The video (after the break) starts off a bit dry, but trust me – it’s worth checking out. The monitor sports a TI MCU, bluetooth chipset, flash socket, multiple signal amps and onboard audio amplification. The PocketPC is showing the real time data stream being delivered via bluetooth.

34 thoughts on “Pervasive Health Monitor (Got Granny?)

  1. This is very cool, and very cleanly done. I wonder what kind of battery life they are getting with the LiPo battery? It’d be neat to see them add a couple more data inputs on the subject and maybe the environment too. With an appropriate program on a cell phone, or maybe an add-on GSM/GPS module I could see this being given to someone with a heart condition to monitor them (send out a update every hour or so) or alert the proper authorities if it detects trouble.

  2. Realistically – as soon as the program detects a problem, it could send an emergency signal immediately. not waiting for the fail-secure hourly messages could save the few precious minutes in the example of a heart attack victim.

  3. i was talking to the guy who made this, the other day. apparently it could easily be adapted for any monitoring application. however if you just want to monitor voltages, you’d be better off hooking an ADC directly to the pocket pc, bluetooth and an onboard accelerometer and SD card would not be necessary. i asked what the battery life was, but didn’t really get an answer yet….

  4. The ECG (heart tracing) is moving too fast, unless your female friend is some sort of Olympic-level athlete. It should run at 25mm/second to conform with standard medical equipment that does the same thing.

  5. I’m a tactical medic. I’ve been looking for something like this for YEARS. Hell, we looked for a way to monitor our guy’s vitals when they were in Level A suits, and we were told it couldn’t be done.
    Wish I had one of these. Someone give that guy a contract

  6. Intel has something similar called the Shimmer . The big advantage of the Shimmer mote is that it has an 802.15.4 radio chip on board (the CC2420 from Chipcon(recently acquired by TI)) as well as the bluetooth module. The CC2420 is much more power effecient for intermote communication and for communicating with a base station that supports it. Check out and for info.

  7. Thanks for all the complements!!

    As for battery life, I’ve not measured current consumption yet, but every part we chose was scrutinized to be as low power-consuming as possible. Some of the components were significantly more expensive than their power-hungry counterparts. The battery in that demo is 450mAh, and our goal is to have the battery last 24 hours – that gives us 40mA of current to play with, which is more than enough given the design (as long as the bluetooth isn’t on the whole time). I need to spend some time measuring currents, but I will do that with the next version that should be ready whenever maxim ships us the audio amps (backorder).

    As for this compared to the Shimmer, the big difference (if I’m not mixing up projects) is that the shimmer needs multiple devices and each one is specific to one task. So you need a lot more hardware, a lot more devices, a lot more complexity, and a need for lots of power as all these devices need to communicate together all the time. Our device’s design concept was that there’s one device located on the chest, with sensors all along the chest. It’s simpler, less points of failure, and no need to turn on the bluetooth that often.

    I’ll put up an easier to find schematic and some details on the hardware, but for now, the firmware is going to be closed till we figure out how this project is going to be funded.. Lots of interest, but no checks in the bank yet.

    We’re also unsure about the firmware for the bluetooth chipset — the company CSR has all kinds of strict controls so we might violate our license if we give that out as it exposes their API.

  8. First… Perhaps I’m reading it wrong, but your 100 ohm resistor as a regulator is a huge power loss and just plain wrong. Look into switching regulation. Linear tech has a lot of easy ones. that’ll be your first and biggest power saving feature add. Also, look into power and sleep modes for the IC’s your using. If your using a SPI Rom to load code, turn it off [or sleep] when your done. RF tranmission is very power hungry. Consider limiting data flow, or packetize the data, rather than ‘stream’ it… and even then, perhaps put your BT asic too sleep between packets… this might sound funny to put it to sleep for only 20 or so mS, but it’ll save power. Don’t use any LEDs. If you absolutely have to, use the micro-power LEDs ~4mA, and only ‘blink’ them once a second or so.

  9. You are reading it wrong. The 100 Ohm resistor is part of a Low-Pass filter to clean the noise on the small signal supply side. As I said in the video, the voltage regulator is on the battery and will be on the device in the next version. It’s a linear regulator; a switching regulator would introduce way too much noise into the system. I’m also aware of sleep states of the MCU and it does turn off when not being used, but thanks for the suggestion. The next version will also have an LED, it’s not micropower, but it’s connected to a PWM output so I can set the duty cycle really low on it, as well as blink it. It’s more there for development purposes, either way. The pulse oximeter does have two LEDs in it, but they will not be on most of the time.

  10. A linear regulator isn’t bad if your battery pack is very close in voltage to what your circuitry needs. although you still have the linear’s drop. People have done neat things using a switching regulator followed by a linear, to get close to the efficiency of a switcher with the regulation quality of a linear.

    I’ve been trying to build something somewhat like this for a long time, although all I want is the EKG/wireless function. This is a lovely project and I hope funding comes through.

  11. this is pretty crazy. good job rez! sell this to the government! I can imagine soldiers wearing monitors like these and field medics (like the person that commented above) just whipping out their phone, connecting to the soldiers ID, and immediately seeing their vitals before they administer care. if you could implement a standby mode, which cut power consumption, until the bluetooth picked up a pairing it could prolong the battery life. I don’t know if that’s plausible, but just a few ideas! good job, and I mean really good job!

  12. this indeed is a great idea and project. as stated before this would be an incredible product that should be produced on mass. I mean added programming and you would have a warning system for elderly that could be located via gps and call EMTs in for service. how much more are we looking for? INCREDIBLE! post up a tutorial and I am def going to build me one.

  13. I don’t think using this as an emergency warning device is such a good idea. What happens if the old lady forgets to turn it off when she removes the electrodes to take a shower? A whole brigade of EMS personnel just shows up anyway? I think this has much better use in the military field, especially if outfitted with a longer range RF transmitter. (or perhaps GPRS or cell net connection)

  14. You can tune very easily for the switching noise. You could switch at a freq far away from what your monitoring, and what your monitoring is a very slow signal compared to any switcher. If you’re thinking it would interfere with the ADCs, you could filter right at the input. I’ve found that I can acheive very low noise with switcher’s with very little real-estate. I’ve designed XM radio reciever’s with bluetooth and audio circuit’s all on the same [very small] PCB with very good results using switchers. Linear’s are easy, switcher’s are scary, But, I truly think if you’re going to achieve a very long battery life, you will eventually have to go this route. In the manufacturing world, every penny counts. And if you can save $2 with a smaller battery, You’ve just made the company $50 million dollars.
    Also, if those two mounting holes are plated, go ahead and have them non-plated as it will interfere with the transmission pattern of the antenna. In addition, to save cost, you could implement a PCB trace antenna. If memory serves, I beleive it’s a .8 inch long trace at 60 mil wide. Google for “Microstrip antenna”.
    Also, most importantly, always put on your PCB’s “Patent Pending” even though it may not be.

  15. I’m wondering what, if any, safety features you incorporated into this? All FDA approved medical devices have to be proven to have little to no risk to the patient (or in the case of more extreme devices, the rewards outweighing the risks.) Is there anything that protects the patient being monitored from the device should something go wrong? I know it’s just a 450mah battery, but if this device is intended to be marketed then it is something that needs to be addressed.

  16. Very nice work!

    One question though: does the software work with smaller screens like that of the ubiquio 501?

    P.s. I`m on holiday in Tunisia at the moment so i can,t read the ppt and the azerty keyboard is a pain in the as… i mean neck.

  17. Just saw the 3rd revision details, a pulseox added to this thing would make it gold for combat medics! If you develop a rugged version of this to meet mil standards then I can’t see why your kids won’t be going to private school.

  18. Hi, this project is awesome. I was wondering if you could send me the details to replicate the project. I wanted to do this same project but instead of using a pocket pc, using it with a gp2x.


Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

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