A $40 Arduino and an iPad make a scooter rig so convincing it passes for the real thing — six months of validation done in hours
Scooter, screen and duct-tape: from hack to habit ft. Unni K. Manikoth & Harish Kumar | Config 2026 · Unni K. Manikoth
22 min total·Actually worth watching closely: ~12 min·3 must-watch clips
- 0:12 – 5:52Listen
Why scooter experiences have to be designed in the real world
In India a scooter is bigger, faster and a lot more emotional — it's how people travel across the city — and around half a million Aethers are on the road. The interface has to hold up in sunlight, rain and at speed, with riders far more distracted than most, so a screen prototype can't validate it.
For a hardware-software product, whether the experience works can only be judged in the real context of use
Mostly spoken setup on the market and the design challenge, no critical visuals — listen to it like a podcast▶ Jump to 0:12 - 5:55 – 9:47Watch
Building a test rig to skip a six-month validation process
A segmented display, like a calculator that also has to do navigation, music and call controls, with physical switches designed alongside the interface. The traditional path meant months of vendor tooling plus two more to bring the display up. Instead: a portable rig running completely outside the scooter, with every real control present, testable on a table or on the road.
The UI architecture depends on the kind of switch you pick, so interface and hardware have to be built and felt as one
The live rig demo starts around 532s — start, mode switch, throttle — you have to see how convincing it is▶ Jump to 5:55 - 9:56 – 14:20Watch
The reveal: iPad + switches + Arduino, then it outgrew the project
It isn't a segmented display, it's an iPad running a Figma prototype, with spare-parts switches and a $40 Arduino converting analog signals into keystrokes — basically a keyboard they built for themselves. Then the rig outgrew the project: the same setup carried into the touchscreen world, and the team started building whatever they wished existed.
Months became hours, and by the time it reached a vendor the concept had become a design spec
618s takes the three parts apart, 736s shows how it extended — both worth pausing on▶ Jump to 9:56 - 14:20 – 16:16Listen
Reinventing voice expression for a scooter
Existing voice visualizations are built for a 10-to-12-inch viewing distance; on a scooter it's at least a foot or two further, so the expression had to be louder, simpler and glanceable. That became the chance to build something uniquely Aether: voice visualized as electrons that move, react and respond.
A unique context of use forces you to drop the generic pattern and invent your own expression
Mainly spoken design reasoning with supporting visuals — following the logic is enough▶ Jump to 14:20 - 16:16 – 20:02Watch
An AI-built exploration tool that became the shipping pipeline
Motion tools are optimized for craft, not exploration, so the team built one with Claude and Gemini: around 60 live parameters over the particle visualization. Adding live voice input and the listening and thinking states turned reviewing visuals into experiencing what the rider would feel, and the tool ended up exporting Lottie straight to code. It's live in India.
A tool built to answer one little question can become the whole pipeline, from exploration to shipping
976s live parameter tuning and 1051s voice interaction are the most information-dense moving visuals — a static description can't replace them▶ Jump to 16:16 - 20:05 – 22:26Listen
The transferable method: see the invisible problem, break it down, step out
The talk lands on a way of thinking: notice the invisible problems everyone accepts — this doesn't fit the timeline, there's a factory pipeline in between — break them down to the smallest possible problem solved in the right order, and don't let your JD define what's possible. Build the things you wish existed.
What they found wasn't a rig and definitely wasn't a tool — it was a way of thinking
A purely spoken close with no demo visuals — good for listening and noting the takeaways▶ Jump to 20:05