Like most ambitious new products, it’s been a few years since we first started seeing tidbits of the Oreka O2 cycling treadmill trainer. Which, is a good point: I’m not even sure what to call this thing.
Officially the company bills it as a ‘bike trainer’. But to the rest of the world it looks like a bike treadmill. There’s no running on it though. So, I’m going to keep ya on your toes and call it all sorts of things in this article. At this point, Oreka has only made one product – so it’s easy enough to even just call it The Oreka Thing.
No matter – as you’ve deduced by now, this is an interactive smart trainer of a bike treadmill. It acts more or less the same as a Wahoo KICKR would, except that it’s a giant treadmill that you place your bike on instead. You might wonder how it’s different than something like the Tacx Magnum bike-treadmill thing? Well, that costs over 10,000EUR, whereas this costs a mere 3,699EUR. You noticed that Euros denomination there, because both are only available in Europe at this time. Probably cause of the size.
Speaking of which – the company shipped one up from their headquarters in Spain back this past fall, and I’ve been poking at it since then. Once I’m done here, they’ll pick it up in the following weeks. Until then, let’s talk boxes.
What’s On The Pallet & Setup:
Your smart trainer at home came in a small microwave sized box? Aww…how adorable!
This beast comes on a full size double-pallet. Not one of those wimpy single pallets you might make a trendy hipster couch from. Nope, duo time here. The kind of thing you move big, expensive, heavy things on.
I enlisted the help of a real-life pallet mover, plus two fake ones: GPLAMA & DesFit to get it where we wanted it:
Mostly, I took photos and they moved things around. That’s called delegation. Later, they made me walk home. Lama rode the pallet mover home:
In any event, after you simply remove the box top, you’re left with the parts:
Putting it together is actually relatively straight forward. The unit comes in basically just a couple pieces:
A) The treadmill part (pretty obvious above)
B) The safety bar (three big tubes above) + bolts
C) The cords (look like electrical cords)
D) Some skewer stuff (looks like bike trainer skewer)
E) The rope/bracket connecting your bike to trainer (pink thing on ground above)
It’s one of those jobs that you’ll look at and assume is an Ikea nightmare, but in reality only takes a few minutes. Even less time when you sucker two friends into doing it.
Essentially, it comes down to three basic things:
A) Plugging in some cables
B) Attaching the handlebar/emergency button
C) Attaching your bike to the doohickey.
The handlebar comes in three pieces to put together. As part of that, you’re also attaching the emergency stop button to the top, and then ensuring the wire is properly run down the handlebar to the front of the Oreka platform. It can be a little finicky to do solo, but with two people it’s pretty easy.
Then you’re connecting the emergency stop button to the inside of the front of the frame. You’ll take off the front of the frame, consider the number of things you shouldn’t touch, find the right spot, and attach it. Close it back up and pretend you didn’t see behind the curtain.
Then, on the back the frame you’ll attach a bracket to your bike. As part of that you’ll need to swap out the quick release skewer on your bike with this new one. I guess it’s not really a quick release skewer since it takes a hex wrench to remove.
This one has these funky little metal pieces off the end, which hold the bracket so you don’t ride off the front of the treadmill. You’ll position it vertically first, and then rotate down, which completes the lock:
It’s not clear one way or the other if you’re supposed to ride this quick release skewer outside, but I did just fine for hours. I figured if it works inside to hold my weight, it should hopefully work outside (like any other trainer skewer).
Once that’s done, then just attach the pink robe system to the back of the rear frame and you’re good to go. It uses a double-lock system, so it won’t go anywhere without pressing the inner button first and then sliding the pun up.
Setup isn’t hard. It’s quite easy actually. With that, let’s dive into how it all works. Oh, wait – plug the power cable in and press the button:
Ok, now we’re read to roll.
How it rides:
The Oreka platform might look like a giant treadmill, except there’s a couple of key differences. First being you can’t run on it. That’s because for the most part it’s the weight of the bike pressing down on the belt, which in turn provides forward movement. Sorta like an unpowered treadmill. Your bike though sits atop rollers under the belt. You can see these here below where your rear and front wheels sit:
Whereas the rest of the surface is simply a hard flat material below the belt. You can step on it, but running on it won’t work (trust me, I tried). There’s no way to get the belt to move without forcing the rollers to move.
The point of all that pink rope at the back isn’t to replicate a kinky 50 Shades of Grey scene, but rather to keep your bike from flying off the front of the treadmill. Without it, you’d go off the front instantly. You’ll have constant pressure on that rope at all times (except when you don’t, more on that in a minute).
The rope though pivots, allowing you to sway left and right, which in turn means that you can move about the treadmill/platform just like you would on rollers.
Which, is a good time as any to note that the Oreka system is very similar to riding rollers. Though, I think there’s two key differences:
A) First, obviously, there’s a giant handlebar on the side which increase confidence (though, it also increases ways things can go horribly wrong)
B) The back rope system of the Oreka minimizes swing of the rear of your bike, so that almost puts into slow motion any potential crashes. I know it sounds
To an experienced roller rider, neither of these will matter much. To an inexperienced roller rider, this will make the transition easier. Make no mistake, you can still fall off.
And it’s partially for that reason I quickly decided to wear a helmet. In my case, the only place I could put this thing was in the concrete lower level. If I fell off of it, my head *will* hill concrete. There’s no two ways about this. So, a helmet was logical. But I didn’t wear it initially. First, let’s start riding.
To get up and going you’ll pair it up to an app of your choice. Zwift, TrainerRoad, The Sufferfest, whatever. I talk about that in the next section.
Then, as you pedal you’ll feel the resistance fade away. At first, you’re providing all of the momentum of the platform. Kinda like trying to take a potato sack down a slide. But after a short bit the platform takes over, and your pedaling effectively keeps things similar to out on the road.
Typically when I talk about ‘road-like feel’ of trainers, I’m primarily talking about acceleration and deceleration. Flywheel momentum effectively. How well does it replicate applying a surge of power, or sprinting. But also – how well does it replicate letting off power. When you surge outdoors you’ll feel that spin-up of acceleration, but also an equally realistic deceleration depending on the terrain.
However, with the Oreka, there’s a second element to consider: Does it actually *feel* like the road. In other words, does it feel like my tires are on the road?
And oddly enough, it actually gets closer than any other indoor system I’ve tried, on that count. Because your bike isn’t statically set on three rollers (like indoor rollers), you’re actually on a belt which sits atop an array of rollers that you slowly drift back and forth over. So you feel that nuance, which roughly tricks your brain into thinking it’s on old but well maintained road.
The second piece – acceleration/deceleration, meaning, how well do you feel the inertia? That part is considerably worse than other indoor systems I’ve tried. The main reason is that as you sprint, you strain forward on the rope, which has a little bit of elasticity in it.
But the split second you stop pushing forward hard, it instantly snaps you back against the rear of the frame:
The reason for this is simple: The belt speed isn’t actually that finely controlled to deal with this. And so at this point the entire electronic belt system is still maintaining speed (if not accelerating), while your momentum on the bike frame has started to deceleration. If the belt was 50-meters long, you’d never notice. But you’ve got approximately 3-6” (~8-15cm) of play here, which at 20-25MPH (upwards of 40KPH), happens instantly. Here’s a short video of it:
In essence, the moment you start sprinting or end sprinting, you feel like a dog yank back on a leash. However, for normal steady-state riding, the sensation is just fine.
One function that caught me off-guard pretty quickly into my first ride with it though was the safety system. As I mentioned above, I started my first ride without a helmet.
However, a few mins into the ride I noticed abnormalities with the power accuracy. So, I glided to a safe dismount speed and popped off the bike, straddling on my bike while standing on the side-rails (just like any normal running treadmill) to double-check some tech stuff. Except to my surprise a rumble in the jungle occurred between my legs. The treadmill belt started to accelerate faster and faster, despite me (and Zwift) being stopped. My bike, being attached to said treadmill belt, was now meandering at high speed (I’d guess 20MPH/30KPH or so) on the belt, skipping a bit without any weight on it. You can see this moment play out the first time for me below:
After getting control of my bike by simply lifting it off the belt I pondered my situation. See, unlike the incredibly expensive Tacx Magnum, this unit doesn’t have any optical sensors along the side to detect if you’ve left the belt. Meaning, if you fall off your bike, there’s no system in place to stop the belt. If you look closely at the Tacx Magnum design, the entirety of the length of both sides of the treadmill are optical sensors to see exactly where the bike is.
This is used both for controlling speed during accelerations, but also as a safety measure in case of a crash. Instantly.
Further though, unlike the Magnum, you’re actually ‘tied’ to the Oreka platform via that rear skewer system. So if you crash your bike onto the platform (such as getting caught in the handlebars) and can’t reach the emergency stop button, I don’t see any positive outcomes.
After my initial surprise, I hit the emergency button and added a helmet. I then tried to repro what occurred, and found I could easily do so. Turns out, the belt will re-accelerate if you don’t come to a complete stop. I discussed this with Oreka, and this is expected.
They noted that they haven’t had any complaints from others once they were aware of how this works. And that’s true, when it comes to stopping, now that I understand how it works – that’s fine for casual stops (weird, but fine).
However, my concern is what occurs in a crash. Being tethered to the platform as a belt accelerates below me unstoppably isn’t super ideal.
I think this problem is actually solvable via some additional safety mechanisms. While what Tacx has designed with their crazy optical sensor sharks with lasers system is enviable, I suspect even just a simple connection to a cadence sensor would resolve the issues here. After all, while outdoors if you stop pedaling then eventually the speed stops too (unless headed downhill). Since the Oreka doesn’t provide forward-drive during downhills from Zwift or other apps at present, there’s no issue with simply having the belt cut-out after perhaps 2-seconds of zero-value cadence data.
I don’t doubt there are reliability issues with that approach, but I think this piece needs re-thinking from a safety standpoint. It is notable that there is a wire that runs up into the rear of the frame. It’s not clear to me what this wire does, though the manual does loosely imply a sensor up there somewhere for something:
Anyway, we’ve covered some of the ride feel pieces above, let’s dig a bit deeper into 3rd party app integration and compatibility.
App Compatibility:
The Oreka platform loosely follows the app compatibility standards and industry norms as you’d expect from a high-end trainer. As you probably know, apps like Zwift, TrainerRoad, SufferFest, Rouvy, Kinomap and many more all support most of these industry standards, making it easy to use whatever app you’d like. If trainers or apps don’t support these standards, then it makes it far more difficult for you as the end user.
I say loosely because while Oreka transmits these standards, it doesn’t appear as compatible with apps in real-life testing as I’d like (primarily on the Bluetooth Smart side). Don’t worry, I’ll try and explain.
The Oreka platforms transmits data on both ANT+ & Bluetooth Smart channels, as well as allowing interactive resistance control across both ANT+ & Bluetooth Smart. By applying resistance control, apps can simulate climbs as well as set specific wattage targets.
The unit supports the following protocols and transmission standards:
ANT+ FE-C (Trainer Control): This is for controlling the trainer via ANT+ from apps and head units (with cadence/power data). Read tons about it here.
ANT+ Power Meter Profile: This broadcasts as a standard ANT+ power meter, with cadence data
Bluetooth Smart Power Meter Profile: This broadcasts as a standard BLE power meter, with cadence data
Bluetooth Smart FTMS (Trainer Control): This allows apps to control the Oreka over Bluetooth Smart (with cadence/power data)
In theory, between all these standards you can basically connect to anything and everything you’d ever want to. Be it a bike computer or watch, or an app – it’ll be supported.
But in practice, it was a bit rougher. Personally, I’m lazy. So I tend to use an Apple TV to run Zwift most of the time, because I don’t have to deal with setting anything up. So I started there:
And yes, it did show up in the menus and pair just fine as a Bluetooth Smart FTMS device. Meaning, Zwift would be able to control the trainer, including things like increasing grade or resistance via Bluetooth Smart. It would also include power numbers and cadence, though at this juncture the cadence values are static (and will be removed).
The challenge through is that in my testing with Bluetooth Smart and Zwift, it would refuse to change the grade. So while it showered the power the Oreka platform, it didn’t change according to the gradient of the terrain in Zwift.
After circling back to Oreka, it sounds like there’s some issue with the implementation of the protocol, which Zwift has confirmed to them. The two companies are hashing it out.
So instead, I switched over to running Zwift on a laptop using ANT+. In that case it leveraged ANT+ FE-C for control instead of Bluetooth Smart FTMS.
And there, it worked just fine. It controlled the incline as I approached hills, but only hills equal to or greater than 4%. Below that, no change in resistance occurred. Sure, it shows it on the screen, but it doesn’t actually change the resistance on the unit.
In talking with Oreka about it this is a limitation they are aware of. 3% and below grades are not replicated. This doesn’t impact how fast you go in Zwift, since that’s purely a function of your power output, but it does significantly impact the realism. This is especially notable in courses with lots of rollers, because what happens is you effectively ride the first portion of a hill 0-3% as if it were flat. But then 4% hits and it’s a jolt, because you basically go from 0 to 4% instantly.
A quick note that while the Oreka platform as an FE-C trainer could support calibration, there’s no calibration offered by the Oreka platform. Don’t worry, we’ll talk about power accuracy later.
For me, in my testing, I use Zwift and TrainerRoad as my two main apps (which are the two main apps I use personally). In the case of Zwift, I used it in regular riding mode (non-workout mode, aka SIM mode) as well as ERG mode (workout mode). Whereas in the case of TrainerRoad I used it in a structured workout mode (ERG mode). I dig into the nuances of these both within the power accuracy section. So let’s switch to that.
Like with Zwift, the Bluetooth Smart control didn’t work on TrainerRoad either. So I switched to ANT+ there (per recommendation of Oreka). You’ll see the trainer enumerated in a fairly similar manner on TrainerRoad as well:
Also, TrainerRoad’s standard-issue tips page on using smart trainers in ERG mode. Ironically, I didn’t think about it at first, but you should actually apply the last one here too with the Oreka: Put it in a small gear up front. I’ll dive into why in the accuracy section.
Once you’ve got things paired up in TrainerRoad, there’s little options there to tweak on the Oreka. You can now use the Oreka platform as a smart trainer whereby it’ll control the resistance of the Oreka platform automatically according to the target wattages in ERG mode:
(It appears Oreka is still using the ‘development’ ANT manufacturer ID, as opposed to their own assigned on. Each manufacturer – from Garmin to Polar to Wahoo to Suunto has a specific ID assigned. This has no impact on anything here, just a minor thing they need to fix.)
Finally, at this point Oreka doesn’t have any consumer facing app available for the Oreka platform. So there’s no further configuration or settings to tweak outside of the standard protocols and 3rd party apps.
Power Accuracy:
I’m going to do you a favor. Not me a favor, but you a favor. I’ll tell you straight up it appears to struggle slightly on accuracy. Just a little, only off by hundreds of watts inaccurate.
That allows you to skip this section. A section I’ll still write mind you, but you get to skip it. Like being the teachers pet or something.
With that, onto accuracy.
Typically I talk about how I test power meter accuracy against multiple products. And I do here as well. I did it against a PowerTap G3 hub, a Quarq DZero Power Meter, a pair of both Garmin Vector 3 and a pair of Favero Assioma pedals. Except it became readily apparent a mere 30 seconds into my first test that the power numbers coming from the Oreka system simply weren’t even ballpark accurate. I’d be pedaling at 250w, and it’d be reading 100w. Here, let me show you:
The purple line is the Favero Assioma power meter pedals, while the teal line down below is the Oreka platform’s power. As one might surmise, that’s a problem. Wondering if something was up, I reached out to Oreka immediately to see what I was doing wrong. Unfortunately, it’s not me.
The Oreka platform only has power estimation, and in this case – it’s not terribly ideal. Oreka says that when the unit is simulating 3% or less (flat roads), they struggle with accuracy substantially. It sounds like this largely has to do with the speed of the system, and that it struggles at higher speeds rather than lower speeds.
Instead, the company recommends using an external power meter with the Oreka platform. You might remember this is the same issue that InsideRide has with their e-motion roller system, it too has less than ideal power accuracy by itself. Both companies recommend the same: Use a power meter. Of course, we’ll get to my thoughts on that in a moment.
Since it took all of a few minutes to establish that by itself the power meter values are less than a random number generator, I connected it up to my Quarq instead. We’ll use Zwift as an example first, but then I’ll show TrainerRoad (to demonstrate two different simulation types). With Zwift, you’ll pair it as a controllable trainer, which automatically pairs it as a power meter. However, after you do that you’ll re-pair the power meter side to the power meter of your own:
Next, it’s worth noting that currently Oreka broadcasts a cadence signal too. This produces no value, so you should actually disconnect that and use something else (probably your power meter is a good idea). Oreka says they’re working to fix/remove that.
With that setup, off I went. Now at this point it’s less about power accuracy and more about power responsiveness. In this setup I’m reliant on the software platform (Zwift or TrainerRoad in my case) to take my power meters input and then look at what the Oreka is saying and figure out how to translate that to a power value that’s offset correctly for the Oreka. In other words, it’s basically applying a constantly changing offset to ‘trick’ the Oreka platform into giving me the correct power.
As such, the above is *NOT* the accuracy of the Oreka, but effectively just the accuracy of the Quarq vs the Favero. Which, is identical.
In the realm of Zwift, that workaround roughly works. It’s not perfect, it’s good enough for most use cases. Note, this doesn’t account for the lack of resistance below grades of 4% however, which I noted earlier on. That piece hasn’t changed, and is still an issue in my opinion. But it does respond just fine to sprints and other efforts largely as I’d expect. Does it feel like a high-end indoor bike? No. But is it mostly acceptable? Sure.
Oh, before I go forward note that with Zwift and TrainerRoad I had to use ANT+ FE-C for all these tests. I did not find the Bluetooth Smart control success on either one (Zwift on Apple TV, or TrainerRoad on a MacBook Pro). ANT+ via both Windows PC and MacBook Pro worked just fine.
Next, what about ERG mode? For that I switched over to TrainerRoad. I like mixing up the platforms to see how different apps handle it. In this case I paired up the Oreka platform as an ANT+ FE-C trainer, and then paired up my Quarq as an ANT+ power meter.
I ensured ‘auto’ was enabled on TrainerRoad, which is the default setting on their power matching functionality. Again, in this case I’m partially reliant on TrainerRoad’s technology here (meaning, how smart it is or isn’t), but also reliant on Oreka’s ability to handle what TrainerRoad throws at it.
And things got off to a rocky start. In fact, so rocky I started and reset three times thinking something was wrong. You can see the reported power at 345w, despite the target being 143w.
What was wrong was that I was in the biggest chainring. This meant that I was at a higher speed than the Oreka platform could handle at the lower wattage of my initial warm-up (roughly 130w). To resolve that, you shift to your smaller chainring up front, and easiest gear in the back. This is actually a funny lesson that I know from testing trainers. But these days it’s usually around response time – not outright wattage floor.
The ‘wattage floor’ is the *minimum wattage* a trainer can output. For most trainers, it’s roughly in the 70-90w ballpark. Sometimes you have to tweak your gearing to get there, but rarely does any trainer I test have to tweak gearing to get to 130w.
What the changing of gearing does is typically increase responsiveness. The faster the trainer (or belt) in this case moves, the longer it takes to respond to ERG mode changes. In any case, once I shifted gears you can see things started to come back to alignment.
So, into the workout I went. At first, things didn’t go too bad as I stepped up in wattage further into the warm-up. There was virtually no delay in the wattage bump, but you can see it’s taking a while to really stabilize after that bump.
This is a core reason why I dislike power matching technology (from any company). It’s essentially a game of high-speed threesome of the app trying to figure out what the trainer and power meter are doing. Someone is almost always left out. It’s much easier if the trainer is just accurate to begin with.
In any event, what I saw was that it would typically take about 5-15 seconds to stabilize power to the target value. This was problematic for a workout (as I did this day) that had 15-second intervals in it, assuming I simply kept my cadence/gearing/speed the same.
And just to be super clear here – this wasn’t some case of my legs being unable to hit these intervals due to lack of strength.
Now you can see some of them were quite close. But that’s mostly because I’ve been doing this long enough to know how to ‘game’ the system. I would over accelerate (via increasing cadence) just after the start of an interval. By doing so I’d trick the threesome into thinking all was well, and I could keep the set point high enough. I did this for some intervals, but then got mentally tired of doing it towards the end.
If you were doing an ERG mode workout with longer (multi-minute) intervals, there’s no massive issue (aside from the wattage floor) with the Oreka in power match. However, if you were doing shorter intervals (sub-30s), then I don’t think it’s viable without deciding to make a game out of it (and thus, likely impacting the goal of your actual workout, which typically involves both a power and cadence target).
Ultimately, the power accuracy here, simply isn’t here. It’s hard to accept that for a 3,800EUR device, no matter the type of device.
[Side note: I have absolutely zero plans of reviewing the Tacx Magnum offering either. I just don’t get it. It’s outside the ballpark of anyone but Sultans and research institutes, and the limited time I’ve spent on it at tradeshows I didn’t really come away impressed with the road-feel bits enough to pique my curiosity. I have no knowledge one way or the other on accuracy, since I haven’t tested it.]Summary:
The Oreka platform aims to fill a specific gap in the market – putting a full bike on a treadmill with the freedom slightly greater than that of rollers, but the smarts of an indoor trainer. Essentially a treadmill version of the InsideRide E-Motion rollers. At a basic level, it achieves that. I can ride my bike on it, and when connected to a 3rd party power meter on your bike, it isn’t significantly different as an experience than the aforementioned rollers.
The challenge I have is that while I think this is a good first attempt, it’s hard to recommend it. For the price, I expect accurate power. Just like I criticized the InsideRide system that’s 1/3rd the price. When $ 500 trainers can give me accurate power, so should things above that. My other concern is around the minimum threshold of 4% incline in order to see resistance changes. And finally, I think a bit more work needs to be done on the safety side of it.
Still, I want to see what Oreka can come up with next. If they can find a way to harness the power they have in the platform and fine tune the software to be more accurate and more compatible, as well as implement some additional safety metrics – I could see it being an option for some. After all, the price isn’t substantially different than some of the smart bikes coming onto the market these days. And hey, if there’s a market for 14,000EUR indoor smart bikes and $ 4,000 consumer running treadmills, there’s certainly one for a 3,700EUR cycling treadmill.
With that – thanks for reading!