Skip to main content
HomeProjectsRetractiles

Retractiles — Tactile Paving That Gets Out of the Way

Published Jul 12, 2025
Updated Jul 30, 2026
2 minutes read

Two users, one pavement

Open with the argument we used in the pitch, because it is the whole project:

Visually impaired user: "I'm so glad the tactile system in Singapore is so comprehensive." Wheelchair user: "No, I hate it. It's so bumpy on a manual wheelchair, and my helper is stressed the whole way."

You have stepped on those yellow studs. They are not decoration — they are how a visually impaired person knows where the platform edge is. They are also, for anyone on a personal mobility aid, a field of speed bumps.

Singapore has around 100,000 m² of tactile paving with roughly 39,000 m² more planned. Physical disability is the most common disability here, about 34.4% of an estimated 45,000 persons with disabilities, and an ageing population means more PMA users every year. So the collision between these two groups gets worse, not better.

Why nobody has fixed it

Not because it is hard. Because the current solution is cheap enough, acceptable enough, and effective enough. Rigid studs get installed one at a time, they meet the standard, and the person who signs the contract is not the person who rides over them.

The other failure is placement. Tactiles routinely run outside the sheltered walkway while able-bodied paving runs under it. That is a design decision that says something about who gets prioritised, and it is worth naming.

Our answer

A modular block whose studs retract into the ground under wheel load and return under their own stiffness. Key properties:

The compliant-mechanism part, honestly

This is the part I presented at the STEAMunity showcase. A flexure is a member that is supposed to bend — you trade range of motion for part count and cost.

Getting the optimal compliant structure takes some skill and mostly luck. Topological optimisation treats the design space as a matrix of pixels and sorts them into compliant and rigid with respect to a desired input–output motion. That is the principled route. We did not have the compute or the time for it, so we iterated on printed geometry and measured what came back.

The stress–strain comparison is the intuition I keep: PLA at room temperature resists stretching and then breaks suddenly; PLA–aluminium stretches, deforms permanently, and breaks later. At higher temperatures plastics act like wet spaghetti — the weakly bonded polymer chains pull apart. Choosing the material is choosing which failure you would rather have.

Where it went

Junior Design Innovator (Advanced) through the SUTD programme, then the startup track: an inclusivity tour at Enabling Village, business modelling for social enterprises, ethical-AI workshops, coaching with Ms Ruo Mei Chua. We showed the prototype at the STEAMunity Public Showcase, where DPM Heng Swee Keat came through, and the public feedback there pushed us to design for mobility-aid users and the visually impaired together rather than sequentially.

We became finalists of the SUTD Create4Good Innovation Fund (S$50k). We did not get the money.

My honest read on that: the story and the drive were real, and the market work was solid — B2B, government agencies and developers as buyers, revenue from pilot installations first, then direct sales, then licensing the retractable-stud technology to tactile manufacturers, then modular repair kits. What we did not have was a finished durability case. A pavement product lives or dies on the ten-thousandth cycle, and we had not got there yet.

We were 19 and under, running a social enterprise for the first time. I would do it again.