A few weeks ago I got in touch with the folks behind Velocomputer, who have been working on a Bluetooth Low Energy Speed/Cadence/Power Meter. The unit is no bigger than your current standard speed/cadence sensor that you mount today to gather those metrics. While here at Eurobike I had the chance to get a closer look at the prototype unit, and try it out.
Note that in the power meter world, there are two types of power meters. Those that are ‘direct force power meters’ (DFPM) and everything else. Direct force power meters measures the explicit force that you exert on strain gauges in the device (i.e. PowerTap, SRM, Quarq, etc…). Whereas non-DFPM devices calculate power mathematically using a combination known, assumed and measured values – to determine or estimate the power exerted (iBike, PowerCal, etc). This particular device falls into the non-DFPM category.
At first glance, the unit looks like any other speed/cadence sensor. And in many ways, the different components serve the same goals. First is the cadence sensor. This is no different than any of the cadence sensors on the market today. Two magnets pass each other in the night (one mounted to the crank and one inside the little pod), and cadence is determined each revolution. Simple and straight forward.
But it’s while gathering the speed data where things get interesting. Like a normal speed sensor, there’s a magnetic bar mounted to the frame of the bike that’s monitoring the magnetic on the spoke. And, like normal speed sensors, there’s also a magnet on the spoke. But where things get different is that instead of the speed sensor waiting for the magnet to pass it each revolution – it’s actually monitoring the position of the magnetic throughout the entire wheel revolution.
You can see above where the magnet is positioned it’s much closer to the wheel hub, versus a tradition speed sensor being positioned closer to the rim of the wheel. The speed sensor itself is where those two zip ties are on the frame to the left. By positioning things close to the hub, and with a strong enough magnet, the unit is able to track the magnet throughout the full rotation. This in turn gives them the ability to produce distance accuracies within .01 meters (1cm, about a third of an inch), and speed accuracy within .01 kilometers per hour. The magnet must be positioned within approximately a 2cm range on the spoke itself – but doesn’t need to be precisely positioned at a exact distance from the hub on every bike. Meaning, there aren’t any installation related calibration issues likely.
In order to illustrate this point, they demonstrated a very slow rotation of the wheel to the exact same position, and then demonstrated going in reverse to ‘undo’ the distance to the exact same point. In order to track the position of the wheel for demo purposes, there’s a small divot/dot on the inside metal rim – just above where the brake is and equal to the valve stem. You can see that start to move and then note when it comes back around again.
As your watching the video, you can see the distance (in the lower right corner) increasing slowly.
Quick hands-on demo video
The power meter determines power by measuring the velocity of the wheel (via the magnet) and then multiplying it by the force. The force is known by simply taking the acceleration (measured) and multiplying it times the mass (entered by you). You enter your weight and the weight of the bike (including things like water bottles, etc…) in the app:
Given the equation, there are scenarios where this breaks down a bit – primarily, windy ones. In outside riding there’s no method to measure the wind – and thus no way to compensate for a headwind or tailwind. The power differences in riding into a 20MPH head wind would be substantial – but even small winds make a big difference. Additionally, if you lose weight over the course of the ride (such as removal/use of water bottles) – you’d have some accuracy issues there (albeit minor).
Now – neither of these are super applicable to track cyclists – which is one of the reasons the product is primarily being tested in that area. Neither is it applicable to folks spending time indoors on trainers.
Which means it might make for a reasonable way to turn any trainer into a $200 power meter. During my test video you can see adjustment to the resistance unit on the trainer being made, and changes in wattage occurring. These wattage values ‘felt appropriate’ to what I was pedaling, but it’s unclear on how accurate they were without a comparison power meter.
[Update: I checked more about how they were calculating the resistance on the trainer with the company. It turns out the trainer power curve was defined in the app (resistance), which means that they’d either have to provide it for each trainer, or it wouldn’t necessarily be accurate. In short, this isn’t ideal for a trainer.]
As demonstrated in the video, the unit will also record slope (as in incline) in 1* increments as well. In the video that when the bike’s front gets picked up, and the phone app shows the values increase and then decrease slowly.
The unit will fit on just about any bike – including mountain bikes, as you can see below showing how the magnet can attach behind the disc brake.
The sensor is powered via a standard coin cell battery and gets about 150 to 200 hours of battery life – depending on which precision settings you use. Note that from a Bluetooth standpoint the power is actually calculated at the head unit (phone) itself, as opposed to be done at the sensor. This means that it won’t likely work across the board with different apps in the future, only with their native apps.
Because the unit uses Bluetooth Low Energy (as opposed to regular Bluetooth), it will only work with Bluetooth 4.0 devices (as it’s a rival to ANT+, it’s not ANT+ compatible). Today the most common phones that support that are the iPhone 4s and the Motorola Razr. The demo here was done on the Motorola Razr (Android). They have the Android app complete, and the iPhone app in beta.
The app records the data to a simple CSV file which can be exported and imported to most training platforms.
The Velocomputer sensor will ship in November, and is available now for pre-order at their site for $200US. Though apparently the first 100 folks get it for $100US. As soon as I have a unit in my hands (hopefully Interbike) I’ll be putting it up against other power meters to see how it stacks up accuracy wise.
As always, if you have any questions – feel free to drop them below and I’ll try and track down the answers.
What do you think about this comnpared to trainerroad? Do you think it’s a better (more economical/accurate) option for those stuck on the trainer for the majority of their rides?
I don’t really understand how it could approximate power on an indoor trainer. Since it uses the acceleration of the whole bike + rider in order to estimate power it can’t work if the bike + rider doesn’t accelerate? And I’m pretty stationary when I’m on my trainer 🙂
Or am I missing something?
I’m guessing it’ll need to know the “power profile” of your trainer, like the Kurt Kinetic power computer???
Maybe it’s a more accurate version of that?
Erik, I would expect it calculates the ‘acceleration’ of the bike by differentiating the velocity of the bike – which is found by the rate at which the wheel rotates i.e. acceleration = change in velocity/change in time.
I don’t suspect it uses an accelerometer to determine the acceleration of the bike.
Mathew: But that doesn’t work in the trainer case. It works on the road because then you know that X kg is subject to Y acceleration and from there you can calculate the power needed. But on a trainer there is no mass that is being accelerated. Instead it’s resistance in the trainer.
similar to the Niterider Rebal power meters which based on speed, weight, altitude, incline/decline
I wonder if it is plausible to put a pitot tube somewhere on the bike, transmit wind-speed data to a cyclo-computer via ANT+ or BT, there it is combined with speed/acceleration/incline to calculate power.
As a bonus you should be able to calibrate zero by simply coasting down a hill.
Yes! it would be a great and cheap idea for a power meter … something like the iBike but as a small ANT+ sensor somewhere in the bike.
Sigma incorporates a similar system on their Rox 9.1 cycle computer.
Are they at Eurobike Ray?
I will take chance at $100 on the early bird deal. It will be better than not having any kind of Power reading.
I’m having trouble understanding how this unit might take hills into account.
I belive the sensor has an Accelerometer built in so it can measure the incline so it should cope well with hills, going up hill is where it should be most accurate as there is less wind effect. It is of course the wind effect that kills all these virtual power devices, they just can not account for a tail wind or a headwind or being on the front of the group or draffting at the back.
Correct, hills are fairly straightforward to take into account with the built-in accellerometer in the small unit (the piece near the crank arm).
The issue is lack of ability to capture wind-impacts (big outdoors), and trainer resistance. Meaning that on a trainer you’d have to manually enter in that resistance value. How it would be done would be difficult. For the trainer I tested at Eurobike, it was static for that specific (Elite) trainer.
This product can’t measure power!
It is extremely misleading to say you can calculate power from only accelerating and gravitational forces. Cycling is all about the drag forces and this product has no way of measuring those. It can estimate them based on an average drag coefficient and rolling resistance but that gives the rider no meaningful feedback.
Not only that but during accelerating phases a large amount of power is due to inertia of the wheels which the sensor can’t measure either. The only way to accurately measure power is to measure pedaling or hub torque.
I don’t think the product can measure cadence either since it only has one rotational sensor.
Combined this makes the product no more useful than my existing smartphone which can measure speed, forward acceleration and grade via accelerometers and GPS.
Not disagreeing with you on the other points.
However, the cadence is still just a traditional cadence magnet. Nothing new there.
Any idea if it can Bluetooth pair with iPhone apps like Cyclemeter or Strava? Any idea if it can export in .TCX or .GPX outputs, in addition to the already published .CSV?
You can definitely export to .TCX (well, .PWX, which is a .TCX format). Plus, the Wahoo app does have direct upload to Strava (if your end goal is getting it there).
I think CycleMeter is supposed to support it, but I’m unclear if they do in this release. They key is supporting the Bluetooth Smart Power Meter Profile (shared by the Wahoo KICKR, Stages Power Meter).
My initial goal is to Bluetooth pair the VeloComputer Smart Sensor with Cyclemeter. Abvio, appdev for Cyclemeter, is not optimistic. I’ve gotten no response to my query to VeloComputer.
I export .TCX from Cyclemeter for import into the Strava web site.
If I can’t pair it with Cyclemeter, a distant second best would be to use the VeloComputer as my primary on-the-road app, and export from it to Strava. That’s a very distant second best, as that eliminates all the nice things Cyclemeter does, as well as undoubtedly significantly delays any ability to use a Wahoo RFLKT.
I’ll update my query to VeloComputer to include a request for status on their support for the Bluetooth Smart Power Meter Profile.
The one challenge you’ll have is that (as of my last discussions with them), the VeloComputer doesn’t actually use the official Bluetooth Smart Power Meter Profile, which makes it somewhat difficult (read: highly unlikely) that 3rd party apps will develop against it.
Will it work with WahooFitness app ?
Unless something has changed, yes, it will. It follows the BLE PM spec, which Wahoo basically authored. And since Wahoo also wrote the Kinetic App and just rebranded it from their own app…
Could you please let me know how much time it will take from Smartphone
1) to find/pair/connect for the first time
2) To connect if paired already , how much time it takes
So have you found out any more about the Velocomputer Smart Sensor? It has been a year since you did the intial review…
Yes, there’s one sitting here on my desk. It just randomly showed up one day. It’s on the list, but isn’t high-priority at the moment given everything else. Sorry!
I can see combining this with a precision cadence sensor to compare the rotational velocity of the cranks with the radial velocity of the wheel and an accelerometer to measure the slope of the ground to get a very accurate estimate of power moment-to-moment. When the cyclist stops pedaling at any moment, the sum of the resistive forces can be calculated instantly by measuring deceleration; Any time the bike slows down without pedaling, a unique velocity curve will be made over which rolling resistance and cda can be calculated, then used as a basis for calculation of power when the cyclist is pedaling.
Combine this with a HRM that measures HRV to get a secondary, independent estimate of power that can be ‘calibrated’ on an actual ride, then used on any trainer to get an estimate of power during the ride. All that’s needed is a decent algorithm to put it all together.
John T, I thought of that same idea, but after further consideration, I don’t think it will work. The rate of deceleration will vary according to what is causing the deceleration. For example, if I (a 200 lbs. person) was producing 300 watts going up a 10% hill and stopped pedaling I would go from about 6mph to 0 mph in about 1-2 seconds. For that same 300 watts on a windless day on the flat ground, I could go about 23 mph. If I stopped pedaling, to do that same 6 mph drop (in this case from 23 mph to 17 mph) would take much longer than 1 second. This is because the effect of gravity is constant, whereas the effect of wind resistance is approx. cubed by each additional mph. So without the computer knowing where the resistance is coming from, it would be impossible to calculate power from deceleration alone unfortunately.
How will this work with the Wahoo Fitness app if the power calculation needs to be done by the app?
No, it doesn’t follow the Bluetooth Smart standards for power meters.
I’ve tried to sort out this info myself with no luck. I sent an email to Velo, but have not heard back from them.
Will this sensor work with the WahooFitness app? Also, does the Velo app work with the Wahoo RFLKT? I’ve already made the investment into the Wahoo products and I’m pretty happy… I just want to add a power meter without going broke! I have a Kinetic Fluid Road machine and I know I can get the InRide system, but I don’t need another HRM and I’d like to not buy more than one sensor.
No, it does not. The Velo doesn’t use any of the standard specs, so it won’t work with any of the standard BLE apps out there today.
Bummer… it looked like the perfect solution! Thanks for the quick response!
Can you recommend a power meter that works with the Wahoo Fitness app… that doesn’t cost a small fortune?
Every other power meter on the market today works with the Wahoo App. The cheapest is the Stages Power Meter at $699 (combo BLE/ANT+).
How about ibike Newton+?
The iBike will transmit over ANT+ to the Wahoo app via the Wahoo dongle. But, I’m not entirely sure why you’d do that since the iBike can display it right there.
Of course, the iBike isn’t a direct force power meter (but neither is the $200 PM seen here).
You would need to GT Firmware Upgrade to get power to be transmitted over Ant+, so $600. It isn’t a direct force power meter and tries to calculate power based on wind speed, angle of the bike and speed so I think many people feel it isn’t as accurate. For example if I’m in my hoods and switch to my drops I just became much more aero dynamic and if I keep the same wattage will be going faster. But how does the ibike take that into account?
The ibike has also accelerometers, so it also takes the incline into the calculation.
Thanks. I’m not at all interested in ANT+ as I have an iPhone 5 and the current dongles are made for the older 30 pin connector. Bluetooth is the way to go. My only issue with BLE is the 2 second lag from the app to the RFLKT. overall, I’m pretty happy with all of the Wahoo products…I would just like to add a reasonably accurate power sensor.
The only reason why you would add the GT key to it would be to add the GPS data from the Wahoo Fitness app (the ibike doesn’t have GPS).
Its how you get the iBike to broadcast the power data, otherwise it keeps the data on the iBike which means it will never connect to the phone.
But if Steve wants a BLE power meter then Stages is the only BLE power meter he could pick at present to my knowledge
If you want one of those velo computers they are free on amazon. There is a coupon code for it. It does not work with any other app, actually it really does now work at all. I tried for hours with a few professional and we could never get the thing to calibrate high enough, and you need to calibrate the thing all the time. Total waste of money and time. I tried emailing the company for help twice and never got any response. Good luck, let me know if you ever get it to work and HOW!
I need an important information because I’m very interested about this sensor!
Does it works with Wahoo fitness app and with Wahoo RFLKT?
Hi, do you know if this unit will work with Suunto ambit3 Sport.