I had an interesting chat recently with Valencell. There are no revelations as such but it was good to hear some of the directions the company and the wider market are heading down.
In no particular order here are some of the things we discussed:
Valencell estimated that optical HR sensors for sport are probably sold in the ratio of something like 80:20 for wrist locations vs. other locations. Valencell are stronger in the market for ear-based sensors and the ratio would be different for them. Surprisingly, more like 50:50.
Whilst the ear is not a ‘convenient’ location for me; for oHR it is a technically good location with lots of blood close to the surface. I had some time with the Jabra Pulse Sport a few years ago and the current Jabra model is the Elite Sport. You can see from the Elite Sport, below, that one of the more obvious market trends is for Bluetooth ear buds that are not connected to each other by a wire.
The global market for secondary sales of earbuds (of all varieties, not included in bundled sales) is in the region of 300 million units. Sure, only a small proportion of people who buy those will be interested in biometrics but it will still be a relatively large market.
As well as Jabra there are several other companies playing a role such as Kuai (below, with the triathlon ear buds), Bose, Sony and Samsung. Naturally the likes of Huawei and Huamei/Xiaomi play a role too.
Valencell have somewhat changed their business model in recent years. They have shifted from licensing to selling prebuilt OHR modules. This has happened relatively recently and relatively quickly; as an example the Suunto Spartan Sport WHR was licensed technology whereas the later Suunto Spartan Trainer WHR and the impending Suunto 3 both use Valencell’s prebuilt BW module.
Valencell have 2 modules
- BW – Benchmark Wrist (Version BW1.2 and the smaller version BW2.0)
- BE – Benchmark Ear
You can imagine the obvious benefits to the likes of Suunto being able to buy an already-working module that requires less R&D; significantly less integration effort; and which deliver a faster time-to-market. You can also add to that some value-added capabilities that are also built-in to the Valencell components via a 3-axis accelerometer as summarised nicely in this image; although it has to be said that the watch vendors don’t often leverage these metrics.
We can surmise that one of the trends is therefore to increase the capability of what the sensor can do. Alongside that are perhaps more obvious trends towards: smaller sizes; improved power consumption from hardware and firmware; and increased accuracy. Perhaps also consider that if two sensors (accelerometer+OHR) can be combined into one sensor then the power consumption of that one unit could well be engineered to be lower than that of 2 separate units.
Other trends include what is being measured with the various light sources in the sensors. Valencell now use Infrared light in the ear and green/yellow on the wrist/arm (like the new Scosche Rhythm 24 for March/April 2018), this can be used to determine HR and HRV/RRi (rest) but attention can also be drawn towards measuring blood pressure as well as sleep, CTL/ATL/TSB, and cardiac events. You might hear stories on Wareable.com of Samsung and Apple attempting to use optical sources to assess blood glucose non-invasively – whilst that may well have very significant sporting and medical benefits, both Valencell and other companies I have talked to doubt that it will be possible in the near future. Here is an interesting read on why non-invasive blood glucose is tricky.
I probably knew, but had forgotten, that Valencell also include a Data Quality Indicator in their oHR signal. Essentially indicating how confident the sensor is that it is sending correct data. It might be an interesting way forwards with a Quality Indicator if the sensor could accept rudimentary user inputs like HRmax. Certainly, in my experience, I would have sometimes liked Suunto’s Valencell sensor to know when it is infeasibly high and the same would be true for other vendor’s sensors too.
Valencell + Bioconnected.comWe already know that companies like Scosche, Suunto and Huawei use Valencell components. There are others but Valencell pointedly advised me to keep a look out this year for companies who might revert back to Valencell technology after having tried it themselves. I can put two and two together here and make 5 but we can probably all have a good guess about this one.
One final point was that ‘OHR testing standards’ might emerge in a few months time from the CTA. Those CTA standards might appear at this URL such standards would enable people like you and me to follow set protocols to see if our HRMs are working properly.
Reader-Powered Content
This content is not sponsored. It’s mostly me behind the labour of love which is this site and I appreciate everyone who follows, subscribes or Buys Me A Coffee ❤️ Alternatively please buy the reviewed product from my partners. Thank you! FTC: Affiliate Disclosure: Links pay commission. As an Amazon Associate, I earn from qualifying purchases.
Interesting, thanks for sharing!
– Any discussion of whether they think Wrist oHR might become reliable one day ?
– Not sure about the “revert back” part as I don’t remember a big player using Valencell initially and then moving to an in-house solution ? I could have seen that for “Mio” (Phillips) with the FR225.
1. 🙂
2. Maybe ‘revert from’ proprietary solution is what i meant.
1. So I guess it’s a “no” for reliability because I can’t see you not asking about it 😉
2. Ok, makes sense
i suspect wrist accuracy will never be fully possible for everyone in every scenario..that would be my definition of reliable and hence the smiley face.
There are other hurdles OHR have to overcome first beyond placement. Darker skin does not play well with OHRM no matter where it is for example. The delay also is a problem. I found the tickr fit (now returned) that was placed on my right bicep just reported slowly to the forerunner 935 (2-5 seconds behind the chest strap). This messed with a plethora of metrics during my usage. That delay is even bigger from a wrist OH.
I am intrigued though in regards to companies that are using in-house OHRM reverting to an external OHRM for their devices going forward. Polar, though considered accurate, is using a weird 6 led sensor, Fitbit’s Purepulse is basically the same thing as it has always been and Garmin’s Elevate….well.
Still having a “better” WOHR for someone like me is akin to them saying “your RHR and ALL-day cal count is going to be much, much more accurate now.” I know full well I will NEVER use it beyond this (other than a long recorded walk…and even then..) so what’s the point of the better sensor, especially when I demand access to the closest to precise information–and that right now is nominally only coming from a chest strap.
no point for you and i. and people like us.
however: more people are not like us than like us.
most people say they want more accuracy.
i’m intrigued by your delays.
It wasn’t unique to the tickr fit, my old scosche rhythm+ did the same thing. It might be because of how fast the sensor is getting the reading/sending the data to the watch (which is how I had it set) but when I compared the two HR the tickr fit was nearly always a few seconds behind the strap.
Also, in my experience…the Tickr fit was NOT as accurate in the reporting as the strap. Close? I’d say 75%. The Tickr fit would be anywhere between 10-20 bpm (when inaccurate) behind what the strap was giving back–and that ended up putting my HR zones really out of place.
I will totally admit, I am overly-critical about accuracy with any sensor-tracked metric. For people that aren’t, sure it’s fine, even great. It’s a whole lot better for activities than the wrist-based but not in the same league (for me) as a chest strap.
P.S- There was one other sticking point I should have mentioned as a problem but in reverse. Women tend to not like using a chest strap for the obvious reason. Having an alternative that doesn’t have to be worn on the chest is a great thing.
I don‘t get the hype about oHR. So much faulty, so much dependencies (on small things), that at the end of the day it‘s not reliable.
HR (in general) has so much variability/fluctuation, that device that tracks HR must be absolutely precise (and accurate). But oHR does not that job.
So, why track HR (via oHR) when the results are questionable?
Makes no sense. At the ende of the day it‘s just a huge market(ing thing) and it‘s all about selling selling selling.
The ONLY good thing I find from the wrist-based OHRM is the ability to track RHR. That’s it. For that alone I want it on my watch. If it’s also giving me a better estimation of calories burned during the day, yes I want it. Beyond that…it’s a gimmick. If it EVER becomes more than what it is, Valencell will probably be one of the leaders in this but not for a long time.
RHR: With really any body movement, a good snug fit and enough blood flow (having warm enough) WOHR can work OK’ish.
Estimating calories: There’s absolutely no relation between HR an burned calories. It varies from person to person and is highly individual; like with the HRmax formulas (e.g. 220-age). If an individual would go to the lab and do some serious testing, then a individual ‘estimator’ (as rule of of thumb) could be generated. But without that it’s a gimmick too.
Sticking to HR zones: There are no HR zones. HR can vary for different many different reasons. HR can be higher/lower by 20 beats and you’d be still in the correct (metabolic!) zone.
Chest strap for the obvious reason: I have observed this too.
+/-20bpm – i agree about variation but i have personally never experienced THAT much in non-illness conditions. personally i’d say +10bpm for ME.
specifically i notice that when i take various things for races like caffeine and beet it and other (legal!!) things. these things DEFINATLEY increase power/FTP, i’d say by about 5% (or maybe its tapering…not totally!!). in fact i experienced EXACTLY that on friday during a near maximal bike session
I do use HR a lot…but then I have to ask how i can use it in race scenarios if the ‘zones’ are 10bpm out
Simialrly If i can increase ftp by 5% in a test then if i translate that down to power zones they are also wrong unless i ue the smae stuff for training too.
At least I have known about these differences for, maybe, 10 years. so i can make some estimated manual adjustment
+-20bpm: I‘ve seen people with it. I‘m personally in the +-15bpm ballpark. Espcially when weather (temperatures) is changing quickly from one to another day.
Power zones: This topic is many folded. I‘m generally not completely sold to the common zoning system. I prefer to look at metabolic ranges. With a powermeter, lactate strips and some spiro (gas analyzing) and of course some thorrow testing (not this 2-3min step tests) it‘s quite easy to find the own metabolic ranges. After a while (read: month to years) I got enough feeling to know at which numbers I can work out for how long and how much I need to fuel. Nowadays I really only need a powermeter and my well ‚calibrated feeling‘.
Szenarios: Training is training, and race is race.
Estimated adjustments: Same for me. At the end a good feeling for the body is key. I personally don‘t track HR any more. Every once a while I get curios, but then I put on a chest strap, but not the quirky oHR.