Summer research project · Grade 3 · Aug 2026 · Selected for the Maker Faire Two-Minute Film Festival

K-Dash

Can I 3D print shoes that make me run faster?

Fully DIY running shoes designed for speed. Kanata scanned his own feet, designed the shoes with AI and 3D CAD, printed them in soft TPU on a home 3D printer, and then ran 400 m to find out whether they made him faster. His dad helped with the scanning, CAD, materials and printing.

Watch the video →Project on GitHub →

Kanata running in his K-Dash 3D-printed shoes on a street — Test run in K-Dash v2

Why make them?

Kanata runs with a running club and wants to beat his friends. After watching a video of someone 3D printing shoes, he thought: if the shoes fit my feet exactly, running should get easier.

Kanata's prediction: a perfect fit + a shape that rolls forward = faster?

Kanata crouching in both finished K-Dash shoes, flashing a peace sign — Finished! Both feet in K-Dash v2

What makes a shoe fast?

Kanata's running club coach explained that fast shoes share three features: a raised toe, a thick sole, and a plate that springs back and pushes the foot forward. To see them for real, Kanata visited a shoe store in Tokyo and compared how different brands use plates and cushioning, with a little help from AI.

Raised toe
Thick sole
Springy plate

Running shoes with raised toes and thick soles on a store shelf — Spotting the features in real shoes

A store display explaining how a running shoe is built — A display showing what is inside a shoe

Turning his feet into 3D data

To make uppers that really fit, Kanata scanned both feet with the Heges app and the iPhone's TrueDepth sensor, at roughly 0.5 mm accuracy. The raw scan also captured the floor and everything around it, so the next job was cutting all of that away until only the feet were left.

The Heges scanning app on a phone showing a foot scan — Scanning with Heges

A raw 3D foot scan that still includes the floor — The raw scan, floor and all

Two cleaned 3D foot models in Autodesk Fusion — Clean foot models, ready for CAD

Designing with AI and 3D CAD

The shoes were designed in Autodesk Fusion. Kanata described the shapes he wanted to AI assistants, which drove Fusion through MCP to build them: a mesh upper fitted to his scanned feet, a raised toe, and a honeycomb sole to save weight. A solid red upper came first, just to check the fit; the lighter mesh version followed.

An AI chat window beside Autodesk Fusion with a 3D foot model — Telling the AI what to build in Fusion

Side view of the K-Dash CAD model showing the raised toe and honeycomb sole — Side view: raised toe and honeycomb sole

Isometric CAD render of the K-Dash shoes — The finished design

Many failed prints

Everything was printed on a FlashForge AD5X at home. Soft filament turned out to be the hard part: regular TPU and PEBA were still a bit stiff, and prints kept failing until Kanata changed one thing at a time - temperature, a 0.6 mm nozzle instead of 0.4 mm, and a simpler filament path. The winner was Siraya Tech Flex TPU Air, a foaming filament that gets softer with more heat, printed at 260 °C for a 70A feel. He printed an upper on its own first to check the fit, then added insoles for comfort.

A table of failed and prototype prints, including a red test upper — Prototypes and failures

Kanata lying on a sofa with his feet up, wearing the red test upper — First fit check: it fits!

The black mesh upper being printed — Printing the mesh upper

Fitting an insole into the printed upper — Adding insoles

A plate hidden in the sole

Carbon plates are hard to make at home, so Kanata 3D-printed a polycarbonate plate instead. Bent hard, it sprang straight back without breaking. To get it inside the sole, the print was paused at the right height, the glued plate was laid in, and the print resumed, embedding the plate in the TPU. Finally the uppers and soles were glued together.

Bending the polycarbonate plate to test its spring-back — The bend test

Placing the plate into the paused sole print — Pause, place the plate

The sole print resumed with the plate embedded inside — Resume: the plate is inside

Kanata gluing a printed upper to a sole — Gluing upper to sole

Weigh it, run it, compare it

Kanata weighed each shoe on the same scale and ran 400 m once in regular shoes and once in K-Dash v2.

Shoe	Weight	400 m
Regular running shoes	204 g	1:56
K-Dash v1 (PEBA + TPU)	411 g	-
K-Dash v2 (TPU Air)	268 g	1:51

143 g lighter

v2 is about 35% lighter than v1 thanks to the new material - still heavier than regular shoes, though.

5 seconds faster

Over 400 m, K-Dash v2 beat the regular shoes by 5 seconds. One run each, so take it as a first result.

K-Dash v2 on a kitchen scale reading 268 g — K-Dash v2: 268 g

Close-up of the K-Dash shoe on Kanata's foot showing the honeycomb sole — Ready to run

Watching the sole in slow motion

Filmed from the side, the honeycomb sole was checked at three moments: touching down, taking the weight, and pushing off. It kept its shape through all three.

Slow-motion side view of the K-Dash sole as the foot touches down — Touching down

Slow-motion side view of the K-Dash sole carrying Kanata's weight — Taking the weight

Slow-motion side view of the K-Dash sole as the foot pushes off — Pushing off

What Kanata learned

Lighter
Changing the material and the sole structure made v2 143 g lighter than v1. Weight is something you can design.
Faster, maybe
K-Dash was 5 seconds faster over 400 m and felt easier to push forward.
But one run is not proof
A single run can't separate the shoes from how he felt that day or the course. That needs more runs.

What's next

Run 3 times each, same course and conditions, in regular shoes and in K-Dash.
Time 1 km as well as 400 m.
Check the sole's shape after running and how the feet feel over a longer distance.

Links

Watch on YouTube →Project on GitHub →Kanata on YouTube →

← Back to Kanata's home Contact →