I had only been a college student for two semesters when I started thinking about what I would build for my final mechanical engineering capstone project. In UC Berkeley's Mechatronics Design course, students first learn the principles of mechanical systems before getting into groups and building almost any project they'd like. Teams have access to various makerspaces and machine shops on and around campus, which contain any tool you could need to build a truly amazing project.
It wasn't until several weeks into the semester that my partner and I had chosen what we would make for the course. After considering ideas like a backyard metal foundry (which would melt aluminum cans to forge ingots), a spill-proof, dynamically stabilized mug of tea (nicknamed StabiliTea), a CNC-controlled DSLR Camera robot (maybe a future project?), and several others, we decided to build a novel electric micro-mobility device: LightDiscs. They were essentially an electric skateboard crossed with a OneWheel. Instead of standing on a board on top of the wheels, however, the rider places their foot on a platform within the rotating wheel.
The objective for this project was to create a novel type of electric micro-mobility device. As far as we could research, nobody has created a motorized side-winding boardless skateboard before, but the idea for this type of movement is not new. Several non-motorized "boardless skateboards" are available commercially, for a relatively inexpensive price.
General operation is as follows. The rider inserts one foot into each wheel, after which a powerful motor drives an internal gear which propels the rider forward.
We theorized that this device would be more stable over uneven terrain compared to a regular skateboard, because each wheel has a larger radius and would thus be more capable of handling bumps. In practice, this design made the LightDiscs significantly harder to ride as well; your feet would no longer be held a fixed distance apart, adding a new risk of accidentally doing the splits if you lose your balance, or if the wheels don't quite at the same speed.
We initially imagined that each wheel would be motorized, with a 1500W brushless DC motor providing torque to each foot individually. We quickly realized that this was not feasible, as this would double the number of required motors and electronic speed controllers, effectively doubling our cost. Instead, we opted to make one wheel an idler, used to steer, while the other is motorized. The discs became easier to put on and more versatile to ride as a result of this design decision. We chose the name LightDiscs, as we envisioned installing LEDs around the wheel to resemble the famous TRON light-cycle style. We didn't end up installing LEDs due to time constraints, but the name LightDiscs stuck.
In this project, I worked on the mechanical design of each wheel assembly, while my project partner Kai worked on the software and control systems. My overall inspiration for the mechanical system within the LightDiscs came from a regular ball bearing. In fact, each wheel is essentially a large ball bearing. Imagine replacing each ball of a standard roller bearing with with miniature bearings which are mounted on shafts connected to the inner race. The inner race is then able to stay flat relative to your foot while the outer race spins around it.
Each skate is subject to both axial and radial forces during regular operation. Radial forces account for the majority of forces we expect, such as when the rider travels linearly or rests their weight on the skate. When turning, however, the skate will experience axial forces as the rider's momentum is corrected. The inner platform of the LightDiscs contains two sets of bearings which permit low-friction rotation while resolving the expected load cases. Axial bearings are clamped to the inner flange of the outer race by structural bolts which hold the device together.
Once I had a mostly-complete design in mind, I began fabricating a prototype out of plywood. I documented the entire build process through short-format vertical videos, which have since reached millions of viewers on various streaming platforms.
We tested the first prototype with the motor we intended to use for the final project:
There were several issues with the prototype which we needed to address before fabricating the final version. Some of these issues arose from the materials we used: the plywood was slightly thinner than a quarter of an inch, which was the thickness of the aluminum that we had designed for. As a result, we had significant tolerance stack-up when layering so many sheets of plywood on top of one another. Further, there were several weak spots in the design of the foot stand, which became obvious when holding the manufactured wheels in your hands.
I modified the design to fix these issues, and began manufacturing the final version of the wheels.
I ran into several issues designing structural supports to fix the foot stand to the inner race of the skates. We wanted to package the electronic speed controller on the device, to avoid the high losses from long lengths of low gauge wire that would have otherwise been required. Instead of designing myself, I decided to outsource this work to a computer, and have generative design create them for me. This was optimal for this use case, as we had complex geometry to avoid and predictable, calculable load cases. In the end, I 3D printed support structures from carbon-fiber reinforced Nylon plastic. Here's a preview of what one of the parts looks like:
After we finished manufacturing all of the parts, I could start assembling them.
Nearly all metal components were segmented into multiple parts to allow closer packing from the sheet material we cut them out from. This fabrication method increased complexity and assembly time, but drastically reduced cost and waste.
Once we had finished manufacturing all of the components, we were ready to put them together:
Finally, the LightDiscs were ready for a test ride:
The LightDiscs are incredibly difficult to ride. In fact, neither Kai nor I were able to ride them for very long, as each wheel introduces four separate degrees of freedom in movement. Standing on them when not moving requires constant adjustments to the rider's balance. Still, they are an incredibly fun, albeit dangerous, way to get around!
We have created a novel type of electric micromobility device, which allows a rider to travel with incredible freedom of mobility, despite a steep learning curve to become comfortable riding them. Despite the great fun we had creating the LightDiscs, there are numerous reasons why a product like this is not commercially viable. One of the largest factors is rider safety. The rider would need to have significant experience and great balance to remain on top of this device for long, but aside from that, there are numerous snagging issues inherent to the foot enclosure. The rotating, sharp components would happily devour your shoelaces, and unexpected obstacles could even break the rider's ankle. The motor is unprotected, and while it does not interfere with your foot in regular operation, it does pose a risk for further snagging. Despite these safety concerns, the LightDiscs are mechanically sound and function exactly the way we intended them to.
There are several changes I would make to these devices if I were to design a follow-up version. The inner foot cavity poses snagging risks for loose shoelaces, so creating a shell to protect the rider's foot as well as keep debris from entering the gearing would be the first improvement. The material the tires are 3D printed from is a high-durometer TPU, which is too firm to give the rider adequate grip. I would replace this material with a softer durometer TPU, and reduce the infill percentage of the tire. Next, I would add a secondary motor to the other foot, to allow simultaneous acceleration from a standstill and improve the riding experience.
At the final undergraduate project showcase, a group of impartial judges awarded LightDiscs the Mechanical Design award in recognition of the principles utilized in creating this project. Later, this project was featured in a video that Buzzfeed made about my story as a creator. That video is linked on this website's homepage, so check that out if you haven't already!
