This new hyper-efficient drivetrain is a cyclist’s dream come true.


Bicycle drivetrains have steadily improved over the years, covering shifting precision and smoothness, friction, noise, weight, and what have you. But one common element in every modern high-performance design — internal gearboxes included — is the roller chain. Those have gotten better as well, but what if you got rid of them entirely? CeramicSpeed’s radical DrivEn concept does just that, and it’s the first ground-up rethink of the bicycle drivetrain in quite some time. But whether the hyper-efficient system makes it past the concept phase is anything but certain.

In search of the elusive 1%

Back in 2017, CeramicSpeed presented itself with a challenge that was internally dubbed as the “Pursuit of the 1% Drivetrain.” As the name suggests, the goal was to create a drivetrain with less than 1% frictional losses, or alternatively, was 99% efficient in transferring rider power from the crankset to the rear wheel.  

CeramicSpeed has come close recently by modifying conventional drivetrains with its array of hyper-optimized products, including the UltraFast Optimized chain, Oversized Pulley Wheel System rear derailleur hop-up kit, and a full suite of low-friction ceramic bearings. But even with that secret-sauce UFO treatment, chains still have a lot of bits that rub against each other.


The core concept of the DrivEn drivetrain is how the machined aluminum teeth interact with the cartridge bearings on the carbon fiber driveshaft. Because the cartridge bearings are freely rotating, there is almost no sliding friction anywhere in the system – at least in theory.

Instead, CeramicSpeed’s new DrivEn drivetrain concept — developed in cooperation with a student research team at the University of Colorado in Boulder, Colorado — uses a novel shaft-drive design that the company says eliminates sliding friction almost entirely.

Taking the place of the conventional chainring and rear sprocket are two “gear rings” with teeth that are oriented roughly 60° from the norm, and sitting in between is a carbon fiber driveshaft that rotates on two low-friction cartridge bearings. But unlike typical shaft drives that use high-friction bevel gears, the ends of the DrivEn shaft are fitted with two circular pinion arrays topped with a series of smaller cartridge bearings. It’s these cartridge bearings that intermesh with the gear ring teeth, and by carefully adjusting the angle and shape of the interface, CeramicSpeed says it can limit system friction almost entirely to the rolling type.


At the rear end is a similar gear-and-bearing interface.

So how much does all of this matter, anyway?

“DrivEn creates 49% less friction than a stock [Shimano Dura-Ace] drivetrain averaged across all gears,” said CeramicSpeed chief technology officer Jason Smith, who founded the once-independent drivetrain friction test facility, Friction Facts. “When that Dura-Ace system was optimized with a CeramicSpeed OSPW and UFO racing chain, DrivEn created 32% less friction. From an efficiency standpoint, DrivEn hits the magic 99% efficiency number at 380W of rider output.”

Still so many questions

DrivEn is very much in the concept phase, and there are countless hurdles standing in the way of the system’s widespread implementation.

For one, DrivEn is currently still only a single-speed prototype; a multi-speed system with nearly 500% of total range (more than what you get on a conventional compact double drivetrain) was envisioned and physically conceptualized for promotional photos, but it doesn’t actually work. But the idea of how it might work sounds promising.


The driveshaft rotates on hybrid ceramic cartridge bearings mounted at either end.

Instead of the truncated cone shape of conventional multi-speed cassettes today, the DrivEn one would be a flat disc. Gear rings of different tooth counts would be nested within each other in the same plane, and by moving the rear driveshaft pinion axially along the shaft, different gear rings could be engaged to change the gear ratio. In addition, changing the bearing counts on the front and rear pinion would essentially be the same as changing chainrings.

Figuring how exactly how that rear pinion would move falls outside of CeramicSpeed’s area of expertise, though, and Smith says the company would likely have to partner with another company to do the heavy lifting. In theory, a small electric motor and wireless receiver would live inside the hollow carbon fiber driveshaft, responding to signals sent by the shift levers.

But even then, there would be big challenges. For example, how would you move such shapes against each other, particularly when they’re moving and under load?

“We talked a lot about this automated system and the brain,” said Alex Rosenberry, the engineering student who managed the project on the University of Colorado side of things (and who now works for CeramicSpeed as a contractor). “The system would have to be pretty intelligent to know how fast the system is moving. It’d have to be pretty smart to know how fast it’s moving, which tooth track to select to make the shift happen. Would you have to back off to shift under high load? We’re not sure yet, but there’s definitely potential to make it work.”


CeramicSpeed has already determined that angling the teeth and bearings can further minimize sliding friction between the two.

There are other issues concerning multi-speed functionality, too, such as the ability to select specific gear ratios. While the flat cassette format provides incredible potential for range, the intermediate steps are still limited by what gear rings will fit within each other. And never mind the rather major hurdle of convincing one of the existing drivetrain players to work with CeramicSpeed on this, or finding someone with enough engineering muscle to pull it off.

On a more basic level, DrivEn obviously won’t work with conventional frames, and anything built specifically to suit would have a long list of unknown issues to tackle to make everything work well. CeramicSpeed’s prototype was based on a modified Cervelo P5 that was made by hand, and any company willing to take a chance on making a DrivEn-compatible frame would be rolling the dice with no certain outcome — never mind the person buying the thing.

And while the claimed friction performance of DrivEn sounds enticing, what about other performance metrics, such as weight, aerodynamics, and durability? How much load would DrivEn be able to handle? What would happen if the system got dirty? And what about compatibility with the matching components?


CeramicSpeed created this one-off sample to demonstrate how the DrivEn system might appear in a more production-ready form. Photo: Allen Krughoff.

That’s certainly a long list of what-ifs, and a seemingly insurmountable one at that. But then again, we’ve seen radical ideas turn into reality before, so who’s to say where this will go.

“How likely is it that it will ever come to fruition?” asked Rosenberry. “I have no idea. That depends on a lot of other things. We certainly think it’s possible, but whether or not it’s going to happen is beyond the scope of what we’re working on.”