Nocturnal Flowers
Product Design, Soft Robotics, Interaction Design
A 3D-printed flower that opens when a shadow falls across it: a light sensor triggers a peristaltic pump to inflate soft petal actuators, revealing the LED inside. The pump is the type used in medical infusion lines, so the bloom has a pulse rather than a sweep.
- Drawing title
- Sheet 04-05
- Organisation
- Role
- Year
- Location
- Sector
- Status
- Built with
Background
INDN 321, Design of Interactive Products. The brief asked for an interaction that responded to research carried out earlier in the course, a cultural probe and interviews into what people found engaging about light. The finding was consistent: people wanted a light that changed in response to them, with some element of surprise, and they preferred materials that read as hand-crafted over plain surfaces.
The concept answers that with a real, slow, largely invisible interaction: the way a sunflower's own circadian rhythm opens its petals toward the sun and closes them as light fades. A lighting installation only works in a dark space, so the idea is built on the inverse of that behaviour: a 3D-printed flower that opens when it detects the absence of direct light, exposing an LED at its centre. In practice, that means a visitor's own shadow falling across a flower is what makes it bloom.
What I built
A 3D-printed flower with soft, hollow petals, wired to a light-dependent resistor and driven by an Arduino and a peristaltic pump.
The pump is the part worth dwelling on. It is the same mechanism used in medical infusion lines, squeezing fluid along a tube by compressing it in sequence, and here it pressurises the petals until they open. That gives the bloom a pulse rather than a servo's clean sweep, so the flower reads as something breathing rather than something actuating, which is most of why the effect works at all. The petals themselves come out of multi-material soft robotics, following research Nicole Hones carried out for her VUW master's on "hydrophytes" that respond physically to changes in their environment.
The concept was a wall panel of a dozen of them. What was actually built is one, in keeping with the brief's own allowance that term, written mid-pandemic, that a working prototype might not be buildable at all.
The problem
The idea and the mechanism are both sound. What lets it down is the finish: the printed petals work, but read as a prototype rather than a resolved object, and the piece needs real refinement of the thing itself rather than more concept development around it.
Outcome
The flower does what it was designed to: cast a shadow over it and it opens, exposing the light inside. That is a working proof of a genuinely interactive lighting concept on a single unit, and the panel of a dozen, where one person's shadow would open a drifting patch of flowers and leave the rest closed, remains the version worth building.

