Earlier this week a well kitted out van drove from Copenhagen to Amsterdam. Upon arriving it unloaded an indoor smart bike that in a few years might be the leading option in indoor smart bikes if you use Zwift. Or, it might fizzle. This post is a first look at a piece of hardware that Zwift believes is so innovative that they’ve licensed the hardware technology inside of it. Ultimately in a deal that allows Zwift to make their own indoor smart bikes.
But like many things in life, the story isn’t as neat and tidy as this just being a ‘Zwift Bike’ prototype. Instead, it’s a bit of a look at the technology of a small company that’s managed to build something both ahead of and behind the indoor training giants. If you want to get up to speed quickly, start with the video. There’s things in there that are hard to clearly cover in text, especially the gaming aspects. Inversely, there’s also a bunch of detail I didn’t fit into the video down here in this post, especially about the complicated Zwift relationship.
VirtuPro is already shipping bikes – and has been for over a year. But, they’re missing the key software integration as of this day that would cause you to be interested in them. It’s why you’ve likely never heard of them. However it’s likely this bike, and to some degree this very announcement, may go down as one of the most important turning points in smart ‘trainer’ history – though maybe not for the reasons you think. First though, as always, let’s get geeky and talk tech.
This isn’t a ‘spin’ bike.
It’s a full-featured, resistance controllable bike. It can simulate grade in an outdoor video, or wattage in erg mode. You can shift virtual gears using clickable shifters on your handlebars, just like a real bike. It can handle thousands of watts of resistance and do so totally silently.
In other words, it’s mostly just like that of the Wattbike Atom or Tacx Neo Smart Bike. Except that you can also change direction of the rider using additional buttons on the handlebars, and that it has a full display built-in.
So let’s start at the front of the VirtuPro Elite indoor bike and work our way back (‘Elite’ is the specific bike model, VirtuPro is the company). First is the display. In the currently shipping models, the Windows powered display is built-in to the unit and included. While we can quibble academically about operating systems, the simple reality is that by going to Windows for this bike you can easily install Zwift, TrainerRoad or many other trainer apps. Or work on Excel while you peddle. Your call.
The display is physically wired to the bike, so there’s no wireless transmission issues that can sometimes be a problem – especially in indoor group environments. Power is provided to the display using one of two cords to plug in, one for the display and another for the bike. As an interesting aside, because the bike is powered and has WiFi in it, with your permission the company can actually remotely troubleshoot the bike in real-time, even getting vehicle-like reports about exactly which components are malfunctioning or not responding.
Now the company says that while this edition includes a display, another model coming either later this year or early next year will ditch the display, allowing them to reduce the cost significantly (more on that in a moment). At which point you’ll connect your own display just like many of the other smart bikes.
Next, we’ve got the handlebars. The handlebars themselves are completely interchangeable to either road or triathlon TT bars, and they’ve got shifting attachments for both types on them. These shifters allow you to change the virtual gears (any virtual gearing you want). There are two lower buttons on both sides, each side allowing you to shift up/down the virtual gearing.
Meanwhile, there’s also singular round buttons a bit higher up on both sides, which act as a way to steer the bike, sorta like a video game controller. This is notable in games that require drafting or otherworldly exploration.
But you know what’s great about all these buttons? They actually make a click. A real-life clickable click of a sound. This may sound trivial, but it’s exceedingly important in my opinion. When you press the shifters on your bike, they click. It’s a physical and audible note to you as a human that your machine has done something at your command and is now executing it. Whereas on the Wattbike for example, there’s no click. It’s like pressing a button on a television remote.
On the flip side, unlike the upcoming Tacx Neo Smart Bike, there’s no slight vibration (technically a sub-second easing of gearing) from between your legs to simulate the moment you shift. The specific design of the Tacx bike and their magnetic resistance scheme allows them to do this (along with simulating cobblestones and such). It’s unlikely that would be possible on the VirtuPro bike.
When it comes to fit flexibility, the unit has all the norms with simple adjustable handles. You can raise/lower the handlebars and seat posts, as well as slide them fore/aft too. All with centimeter markers. Again, the norms.
Down on the pedal and crank front, you can add/swap in your own pedals, and the unit comes in crank length options of 170/172.5/175mm. Unlike the Tacx bike, you can’t change crank lengths on the fly.
On the physical aspects of the bike, it’s got a 10kg flywheel. That’s about half that of the Elite Fuoripista bike, a hair more than the Wattbike Atom, and kinda-less than Tacx’s bike. The Tacx bike is complicated because they use a virtual flywheel, so it’s messy to compare. However, as Tacx has demonstrated with their bike and the Atom, it’s not about the size but how you use it. Even Elite has shown it. Through different routing and belt configurations you can emulate much larger flywheel sizes (like a pulley system). The same is true here too, but I’ll talk about that in the software section.
That flywheel – and the entire bike in total is completely silent. It makes absolutely zero noise whatsoever. The only thing you’ll hear is my breathing if you watch the video, and that’s the only thing you hear in real-life as well.
The internals of the system is driven by a stepper motor that can increment in some 6,300 steps across its 45mm length. What’s most interesting though is how it resets and recalibrates itself automatically after each session. The company says that how they do this is one of the specific and unique hardware features that Zwift is interested in.
Oh, and before we get into the electronics – the exterior case can be customized. I noted my displeasure for the glossy finish, and they said going to a matte-black is trivial. Hopefully also trivial is going to just a single elegantly placed logo.
But there is a gap here in today’s bikes – and that’s ANT+ & Bluetooth Smart compatibility over the standards of ANT+ FE-C and Bluetooth Smart FTMS. These are considered baseline for any trainer or bike today, and not having them makes the bike dead in the water to many. So even despite Zwift’s investment here, as of August 10th, 2018, it actually doesn’t work with Zwift.
VirtuPro says that’s set to change soon. The unit has a chipset capable of Bluetooth Smart, and they could potentially add Bluetooth Smart FTMS compatibility to today’s bike, making it instantly compatible with most apps (or, they could do some creative stuff since it’s wired to Windows, to emulate devices). However, their plan is by end of year to add in both ANT+ & Bluetooth Smart to the device via a small update to the communications board on the bike. Below, you can see some of the behind the scenes bits of the bike where a support technician can run diagnostics – remotely or in person.
It goes without saying, that’s super important to get widespread adoption of this bike. But it doesn’t sound like it’s something the company is worried about. You may be wondering how could a company skip something like this? Well, the reasoning is simple: It wasn’t their target market. Instead, indoor gyms and studios were (and are). In that setting, people aren’t using Zwift and TrainerRoad. They’re using group applications that are mostly proprietary. And in line with that, they do support Polar’s legacy analog heart rate sensors (as those were/are still popular in gym equipment). The company is looking into whether it could actually just insert a USB ANT+ adapter into the platform as an interim step.
Of course, times are a changin’ – and the company is realizing it’s sitting on a goldmine of an indoor smart bike that effectively just needs a $3 chipset added to it to make it not just a competitor – but a potential leader in the indoor smart bike world. Oh, and Zwift investing in them might have changed their minds too.
I get it, you don’t care about their software. You just wanna ride one of the 20 or so apps out on the market today.
But, ignore for a second the user interface of the software and focus on the underlying concepts. These are critical to understanding why Zwift has licensed their hardware and the concepts behind it.
In some ways, the overarching software isn’t that unlike either the Tacx or Elite suites. Slightly dated in appearance, but fairly full featured under the covers. In their app you can ride real-world outdoor famed cycling routes, as well as train against structured workouts that you can create online using a workout creator. You can also do an FTP test and race against others, real or virtual.
And it’s that racing against others that exposes the elements Zwift and others in the industry are interested in, specifically around drafting. See, when you load-up an outdoor route you can add other riders on the course. These can be simulations of professional riders, or they can be your friends. You can even change categories of riders (the CAT system effectively).
You can ultimately place a large group of riders on the start line…for a race. I’m the upper left rider in red with the little halo around him.
As you start riding, you can now move your specific rider left and right on the roadway using the upper buttons on your handlebars, just like a video game controller. Your powered effort, of course, allows you to sprint ahead or let the peloton catch up to you. Gearing is controlled using the buttons as well, and shown via a virtual gear display at the bottom (53/11 in below photo, 34/32 in above photo). The fact that this bike is a triple extends from their current market focus, but changing the gearing is merely a software setting that would take a few seconds.
With that movement left and right, you’ve now got legit drafting. Not just drafting behind a single rider, but their platform replicates crosswinds as well. So the idea that you can create complex formations as a team and you have to get into the exact position to feel that draft effect (they detail that much more in their video here). And most critically, that draft effect is near instant – just like in real life. You can see in the video how quickly the wattage backs off when I get into the draft zone of someone, and inversely, how quickly I feel the pain of the wind when I pop out.
And when would I pop out? Perhaps to do a sprint – which is the second element of what Zwift is interested in, how the system responds to sprinting. Specifically that not only is gearing involved, but the reaction time of the system is fast enough that it feels more realistic. Since I didn’t have a Tacx bike next to it to compare side by side, I can’t say whether one felt absolutely better than the other. But it certainly felt good and realistic, especially when I sprinted over a slight riser of a hiller in the roadway. The resistance fell away naturally and it became more difficult to sprint downhill, it felt super clean.
But from the company’s perspective, it’s the next bit of tech that’s even more appealing to the indoor realm: Coming back from a pause.
Typically if you had put yourself in a ‘hard gear’ and then paused the game for whatever reason, it can be difficult to start pedaling again. The ‘trainer spiral of death’ as its often called, where it feels like going from 0RPM to 95RPM is like pushing a semi-truck. But as you can see in the video, it’s almost effortless (just like on a real-bike outside standing on the side of the road), and the system quickly ramps back to respond to your pedaling/gearing over the next 1-2 seconds.
That’s compared to a trainer where when you stop you’re sometimes left in a weird resistance state that takes some effort to get back out of.
Finally, as noted earlier – the system re-calibrates itself constantly after every session, and in fact, you’ll see that on the screen:
Interestingly, the company notes that they’ve got about a +/- 3% accuracy on the system, which they’ve tested up to 2,500w. After that point, the test system to test the trainers can’t produce any more watts. Interestingly, other trainer companies in the segment have told me the same as well. +/- 3% is perhaps a bit outside the realm of indoor cycling bikes at that price point, where I’d expect +/- 1.5 to 2%. And in fact, I did feel like when I pushed 800w in the sprint for a sustained period it felt a bit easier than it should have. I didn’t have another power meter on this bike as it was only there a couple hours, but down the road if I review one that’d certainly be a key part of it.
It doesn’t sound though like the company has focused a ton of resources on tweaking the accuracy algorithms to date, largely because for their existing market (gyms and cycling studios) there hasn’t been demand to get from +/- 3% to +/- 2%. I think that’s logical given their model to date, but of course, as they shift focal areas it’ll become critical in competing with well established accurate entrants.
So how does Zwift fit into all of this? To say it’s complicated would be an understatement.
The backstory is filled with various major indoor training companies that have been trying to partner or license VirtuPro, such as TechnoGym and Life Fitness (Indoor Cycling Group). It’s also got companies in China looking at leveraging the company for products in markets there too.
There’s a lot of people’s fingers in this pot, and the deal with Zwift isn’t actually exclusive. In fact, after the company left the DCR Studio/Cave on Wednesday they packed up the truck to meet with another indoor training company the next day looking to create a near-term partnership.
All this while the VirtuPro is actively selling the bike to consumers and studios for the last year, and building out their dealer network. Some 80 retailers/dealers are coming online in the next 60 days in Europe, with contracts signed and ready to start delivering bikes this fall via their stores.
On the Zwift front, CEO Eric Min met with VirtuPro last October in Europe and tried the bike for the first time. Almost immediately after the meeting Eric called up VirtuPro with an offer to invest significant sums of money into the company. This led to some internal disagreement within VirtuPro as to the direction of the company (I’m using the term ‘disagreement’ as an umbrella for ‘giant shit-show’).
Ultimately though, Zwift and VirtuPro entered into a licensing agreement that allows Zwift to build their own bike using technology from VirtuPro. The specific technology includes the underlying acceleration and resistance motor pieces, but could be expanded as the two companies see fit. The resultant of which is a royalty agreement in place that’s based on per unit shipments (meaning, as Zwift ships bikes, VirtuPro gets paid for each bike shipped). Additionally, Zwift was sent a bike as well to their headquarters to ostensibly work on integration with it.
Of course, how this plays out remains to be seen. It’s no secret inside or outside the industry that Zwift has heavily pressured trainer companies to build indoor bikes. Zwift sees indoor smart bikes as critical to their expansion plans, akin to how Peloton has thrived with a similar model. Thus why you’ve now got Elite, Tacx, Wattbike, and Bkool all shipping or developing bikes (along with others).
Many know that Zwift often feels like the indoor trainer industry is moving too slowly for Zwift’s expansion plans. And certainly, Zwift is no stranger to hardware focused efforts. You’ll remember in June they acquired Milestone on the running side, to ideally expand Zwift more rapidly there too.
But all this doesn’t mean Zwift is building a bike today, or set to announce one tomorrow. In fact, all indications are that Zwift hasn’t done anything yet with the bike sent to them. Certainly, it’s not integrated into Zwift yet. Zwift did pay upfront as part of the licensing agreement, so it’s not wholly a royalty based one. Thus they have some incentive to utilize that already placed investment. But – it can also be just that: An investment they can leverage down the road, be it one year or three years.
Oh – and in case you were wondering – VirtuPro and VirtuGO aren’t the same entity. But, they are related and there is investor crossover between them.
As I said, it’s complicated (and that’s the half of it).
This story has two tracks to it. One involving VirtuPro as a company making a smart indoor bike just like any other, with the other track being the entire Zwift investment aspect.
For the first train, that’s already left the station with the company already shipping units, albeit to a slightly different market. Once the company can complete their ANT+ & Bluetooth Smart integration this fall, they’ll have a very capable bike that’s quite marketable. Not only that, the company aims to reduce the price to between 1,500EUR and 1,700EUR (excl VAT) by roughly the end of the year (from the current 2,900EUR excl VAT). This is for the variant without a display. Of course it’s not just the removal of a display that reduces the price that much, but a significant shift in manufacturing volume to match the interest they’ve received in the past few months. The company says that aside from the removal of the display, all other components are identical the current bike – just bought now in massive volumes.
Meanwhile, the second train involving Zwift as an investor/partner/customer appears to be hanging out at the station. The doors are open and Zwift has paid for its ticket, but the best I or VirtuPro can tell, that train hasn’t moved anywhere since signing of the contracts. Alternatively, Zwift could have a hardware division already quietly working on a bike and doing so in the utmost secrecy. We don’t know.
As for me reviewing the product in a more formal review setting – that won’t happen until they adopt industry standards, which in this case means adding in ANT+ & Bluetooth Smart. Once they do that that this year (or at least emulate it internally), I’m definitely open to reviewing it like other bikes. Else, I don’t see them by themselves being a big enough player to matter in this scene without those elements.
But for now I’m reasonably impressed. I’m impressed with the underlying concepts of their software and the hardware potential it has and looking forward to seeing it back in the DCR Cave down the road for a deeper dive.
With that – thanks for reading!