Open source Arduino pulse sensor looking for funding | Hacker News Reader

Open source Arduino pulse sensor looking for funding

kickstarter.com

35 pointswebmonkeyuk10 comments

Note that there is a wireless standard for interfacing with biosensors called ANT+ (www.thisisant.com). You can get a USB stick for $30, and a heart-rate strap for $35-40. Most Sony Ericsson Android phones come with ANT+ integrated as well.

So I'm not sure what this adds to the what is already commercially available other than Arduino integration and maybe the fact it's slightly cheaper?

Cheaper is good. Hack-ability is good.

Tiny is good. This thing is minuscule by comparison.

> Cheaper is good.

ANT+ stick is $30, about the same as a 'duino. No idea about the price point of the sensor itself, but maybe $15? So you save $20, lose the wireless connection, and gain a painful earlobe. I'm not saying it's a useless concept, but I think it's applicability is very limited indeed.

> Hack-ability is good.

What is more hackable about this versus ANT+?

heart rate straps like the polar or garmin attach to the chest and require skin contact. This is not always desirable if you are, say, sharing the sensor, or engaging in casual (not athletic) use. Your BOM above does not account for the heart-rate strap signal receiver. Add $15 for the Polar Heart Rate module at Sparkfun, or $60 if you want it on a PCB. Plus the engineering and lashup time. (full disclosure: I'm one of the Pulse Sensor Creators) The Pulse Sensor outputs analog waveform based on blood oxygen content with some people the waveform shows systolic and diastolic components. Heart straps will only send out one ping per beat. This Pulse Sensor may have some limitations. It may get noisy if you're being really athletic, for example, and it might not like being drenched in sweat. But for that it's applicability is very unlimited indeed. Specially if it's coupled with ANT+.

> heart rate straps like the polar or garmin attach to the chest

FWIW, Garmin also makes watches with built-in hart rate sensors and ANT+ integration, although these are rather pricey ($100 including USB stick).

> and require skin contact. This is not always desirable if you are, say, sharing the sensor, or engaging in casual (not athletic) use.

Doesn't your sensor also require skin contact?

> Your BOM above does not account for the heart-rate strap signal receiver.

The USB stick I mentioned is $30 from Amazon. The strap is $35, also from Amazon (prices are similar for co.uk). This is assuming a wireless 2.4GHz heart rate -> PC interface without a 'duino in the loop.

> Add $15 for the Polar Heart Rate module at Sparkfun, or $60 if you want it on a PCB. Plus the engineering and lashup time.

I'm not talking about Sparkfun, I'm talking about off-the-shelf stuff you can buy from Amazon.

> (full disclosure: I'm one of the Pulse Sensor Creators)

I thought as much :)

> the Pulse Sensor outputs analog waveform based on blood oxygen content with some people the waveform shows systolic and diastolic components.

What kind of applications do you have in mind for this? I doubt you'll want to measure blood pressure with this thing, it's wildly inaccurate and if I understand correctly, blood pressure varies throughout the body. It also doesn't really change over short periods of time?

> Heart straps will only send out one ping per beat.

Like I said above, I don't think this sensor is doing much more than that, but I might be wrong. I mean, it's a neat way of doing it for sure, but I don't think you should try to sell it as anything more than a low-cost heart-rate sensor, and maybe a trendy earring :)

> This Pulse Sensor may have some limitations. It may get noisy if you're being really athletic, for example, and it might not like being drenched in sweat.

I believe you guys designed it for MMI, so I don't really see much of a problem in that. Besides, everybody knows nerds don't like to work out anyway.

> But for that it's applicability is very unlimited indeed. Specially if it's coupled with ANT+.

I think your sensor is useful for hobby projects because of its low cost and simplicity, and because it can easily interface with 'duino. I could certainly see it being sold on SF or a similar site. I can also see it used in installations where a user inserts his/her finger into a slot to influence the visuals for example.

> Doesn't your sensor also require skin contact? Yup, but it doesn't need to be strapped tightly to the user's chest. A near impossibility if you are sharing with nerds ;)

Your suspicion of our audience is correct. We don't want to compete with the overly saturated excersize-pulse-rate gizmo market. We see every year our students at Parsons Design + Technology department curious about using heart rate data in a project. We hope this will appeal to entry level and seasoned hackers who want an affordable way to play with biofeedback. The answer to 'what kind of applications' is to look out the window and wonder.

The data that you get from the sensor is variable voltage based on (relative) oxygen level of blood. In the video on kickstarter site, at about 1:20, the waveform you see on the computer screen is raw analog data from the sensor. It is passed directly from an arduino into a processing sketch, which serves as a visualizer. Just thought of an application: Use the fluctuating waveform from the sensor to modulate a tone, say, or make beats.

> Just thought of an application: Use the fluctuating waveform from the sensor to modulate a tone, say, or make beats.

Fair enough, that's pretty cool I guess.

If they used a red/IR LED pair instead of a single green LED they could sense oxygen concentration as well as heart rate. The absorption spectrum for oxygen-enriched blood is different in the two wavelengths. Toggle between the two LEDs fast enough and you'll only need a single light sensor, too.

Yes, blood oxy is easily derived that way, but we were not interested in making a device that had that level of medical accuracy and application. (maybe in future version...?) We just wanted a simple heartbeat sensor with simple interface for hackers. That said, you can get relative blood oxy concentration. A couple of deep breaths will shift the waveform up ~20 points on our visualizer, for example.(I'm a co-creator of Pulse Sensor)

Reply on news.ycombinator.com