If it didn’t work back then, why should it work now? A fair question when it comes to shaft drives and the Ceramic Speed ‘Driven’ prototype drivetrain which has attracted a lot of hype in the last two years. Since 1880 when the first shaft-drive appeared on the Orbicycle (tricycle) by Thomas Moore the subsequent rush of inventions of shaft-driven bicycles in America and the UK in the 1890’s, various bicycle shaft drive drivetrains have been pitched throughout history.
One of the most well known modern iterations is the Biomega Copenhagen city bike by Jens Martin Skibsted which is iconic although the drive was never really that good, because of the inherent efficiency deficits of the drive.

In July 2018, Ceramic Speed made a bold new announcement quoting Jason Smith, the project CTO, “CeramicSpeed has proudly accomplished what many have said couldn’t be done. We achieved a 99% efficient multi-speed drivetrain while eliminating the chain and complex rear derailleur.”

This was enough to launch a media frenzy of coverage which fuelled speculation and debates in online cycling forums. But it was first at Eurobike in 2019 that a ‘shiftable’ prototype of the drive train was presented. I was not chanced with good fortune while attempting to see it first-hand, each time I visited the stand, the demo-version was broken – it was still very delicate. Jason Smith was forthcoming and suggested that in a mere six months, all of the technical challenges for the Driven concept will be resolved.
SIX MONTHS! really?
Despite healthy scepticism, I believe in encouraging ideas and innovation. Even if a first generation new product release is unaffordable, when a product is good then the trickle-down principle means that it should eventually become available to everyone at a more affordable price.
I invited Jason Smith from Ceramic Speed to address the technical aspects of the drivetrain while consciously ignoring all of the marketing and hype. One assumption that is made is that readers are somewhat familiar with the drivetrain, incase you need a primer, skim-read the Ceramic Speed discussion in the Australian Cycling Forums. Speaking with Bicycles Network Australia, Smith also shares hist outlook and prediction when it could be market ready.
Christopher Jones: Shaft drives have not yet been able to establish themselves on bicycles but Ceramic Speed are promoting it as a more efficient drivetrain solution. Has something changed, are there new ideas or approaches that now make it viable?

Jason Smith: Yes, something has definitely changed with the Driven system. Traditional bicycle shaft drives use bevel gears to transfer power. Bevel gears use fixed teeth interacting with fixed teeth. While bevel gears are functional at transferring power, these gear systems are not very efficient. The fixed teeth slide against each other during meshing. Even with the teeth lubricated, this sliding friction is substantial, rendering a traditional bicycle shaft drives at about 90% efficiency.
The Driven system uses ceramic ball bearings meshing with the fixed teeth to transfer power. Each bearing on the front and rear pinions engages, rolls through and disengages a tooth trough. Because rider power transfer occurs through bearings, the Driven system creates only rolling friction. Rolling friction accounts for a small fraction of energy losses when compared to sliding friction. Think bushings (sliding) vs bearings (rolling). With the bearing-style drivetrain, Driven has achieved 99.2% efficiency when lab tested. In other words, only 0.8% of rider power is lost through the drivetrain.
Even though many compare Driven to a traditional shaft drivetrain, it is not really comparable. Yes, Driven uses a shaft, but the power transfer mechanisms are completely different, and needless to say, much more efficient.
Christopher Jones: In some high powered motorcycles, a shaft drive with spiral bevel gears (connecting with the rear wheel) are used however the higher weight and lower efficiency means that they are not adopted for low powered motorcycles. The Ceramic Speed Driven solution doesn’t use the bevel gears (and promotes drivetrain efficiency), but is there a weight penalty over conventional shifting systems?
Jason Smith: We measured the weights of a di2 drivetrain vs the projected weight of the Driven system. We feel the Driven system will be 17% lighter, so probably fair to say anywhere between 15 and 20% lighter. The weight savings comes from the removal of the chain (replaced with a much lighter hollow carbon fiber tube), removal of the wide cassette (replaced with a flat, lighter, single-plane rear cog) and removal of the rear derailleur mechanism and RD cage (but we’d keep the existing battery, motor, and wireless electronics, which would move into the hollow shaft). Really, most of the big gains in weight savings are due to the removal of the chain and use of a flat rear cog.
Christopher Jones: In a conventional chain, there is a much greater area of engagement (where a chain touches a cog) compared with the ceramic speed drivetrain where one or one-and-a-half bearings are in contact with the cassette at any time. On the one hand you may suggest that this reduces the resistance, on the other hand, the stresses and power transfer is centered on a much smaller area. Would this increase wear and tear or the failure rate and how is this overcome?
Jason Smith: Correct, power transfer point forces are greater with the Driven system. We’d have two-bearing “equivalents” sharing the load at all times (this could be actually two bearings fully engaged, or ½ force disengaging, 1 force engaged, and another ½ force engaging, which would always equal 2. Yet, it is important to point out, with a chain drive, even though the total tension on the chain might be distributed across several links as the chain drops into several teeth around a ring or cog, the FULL and TOTAL tension of the chain is ALWAYS seen by the ‘next’ link articulating into the ring. Given this fact, the full chain tension is being transferred through one point, which is the inner workings of a single chain link, as that single link bends to accommodate the articulation of the ring (or cog). So when comparing Driven to Chain, the ‘transfer’ forces are higher with driven, yet conversely, the forces seen at any single point in the drivetrain are higher with a chain.

Christopher Jones: I could imagine the cassette deforming and the pinion with bearings skipping the teeth on the cassette or even exploding under massive loads. How will high-load scenarios be resolved?
Jason Smith: By strengthening the present design, and considering new designs, i.e. we’ve stiffened the front ring and rear cog on our rideable Driven bike. We believe we’re hitting about 350 watts peak with that drivetrain right now. Granted, this is a first step. This was done with simple strengthening practices given the present design.
Then we’re also looking at completely new designs, such as, for example, full integration of the front ring into the crank. And, for example, for the rear cog, picture this- We remove the freehub body and create a single piece rear hub, with the rear cog strongly affixed to the hub, and the spokes extend from the radius of the rear cog, and picture supportive trusswork behind the cog. The freehub ratchet could reside in one of the pinions, completely removed from the hub area. What I am getting at is that the new drivetrain, rather than being limited in power, can actually open up new opportunities for a stronger rear hub area and stronger wheels, given the rear cog can now become a structural member, and the freehub functionality can be removed from the hub. To specifically answer you question, we don’t feel strength will be an issue as development continues.

Christopher Jones: Related to torque, the forces from the cranks to the shaft and then from the shaft (pinion) and cassette change direction. What needs to happen to the parts of the drivetrain and also to the bike to prevent flex. Will the bike frame, rear wheel and bottom bracket area need to redesigned?
Jason Smith: Yes. First off, you are correct with the opposing forces of each pinion. And yes, the drive-side chainstay area of the frame will have to be altered to accommodate driven. The prototype bikes all have had their chainstays elevated to accommodate the driveshaft. Yet, when this system goes into production, it might be a completely different solution. Say, the driveshaft resides within the chainstay, or the chain stay partially surrounds the driveshaft, or the chainstay resides behind the shaft and acts as the back of a contamination cover in addition to structural. Regarding those forces, a present chain drive system induces compressive forces into the entire length of the chainstay. Driven induces separate lateral forces at the point of the pinions (as you mention, BB and rear hub area). They are not huge forces, and it should be relatively easy to design the lower portion of the rear of the frame to support these forces.

Christopher Jones: When we spoke at Eurobike, you suggested that in 6 months, the remaining technical challenges will be resolved… what are the remaining challenges and does this mean that in 6 months, the Ceramic Speed drivetrain would be ready for production? If Ceramic Speed attracted the right partners to put it into production, when would it realistically be available for consumers to purchase.
Jason Smith: It’s now 6 weeks post-Eurobike and we are presently testing our rideable+shiftable prototype in Boulder Colorado. We’re making swift progress. The next step will be to mitigate the vibration in the system. This will take extensive CAD modeling and frankly, a lot of trial and error, to get the tooth shape and meshing perfectly smooth for all of the gears. We will then work on frame stiffening and overall drivetrain strengthening, possible concurrently with the vibration mitigation. However, this will still be considered the prototyping and development stage.
We don’t feel Driven will be ready for production in 6 months. We might have the challenges sorted, and in 6 months, start looking at the production process. It’s quite costly and time consuming to take a prototype into production, so this might be a year or so. (on a side note, we’re in discussions with several potential partners to share development and production resources). All in all, maybe two years until a release. Fingers crossed for a 2022 model of a publicly available Driven bike.
More Information:
Ceramic Speed Driven Project
photos: Supplied by Ceramic Speed and Bicycles Network Australia