Summary
The research team knew what experiments they wanted to run. My job: Turn that into a buildable system spec, then ship it.
Context
Construction-tech startup · concrete 3D printing
The Problem
Industrial printers: Built for construction sites, priced out of labs.
Desktop printers: Clogging, abrasion, inconsistent flow on construction pastes.
What I Worked On
A research-focused concrete 3D printer.
My Role
Owned requirements through testing, present wherever decisions got made.
Outcomes
Problem Context
Output scale vs. Learning speed
Concrete printers optimize for output scale. Research teams need repeatability, fast iteration, and easy maintenance.
Production machine → Research instrument
System Architecture
Motion System
Cartesian motion: Predictable positioning, simple to maintain.
Material Handling
Batch storage and delivery, built to resist clogging.
Extrusion System
Mechanically driven, interchangeable parts for dense materials.
Key Design Decisions
Four intentional tradeoffs.
Coupled system → Modular subsystems
Independently serviced and upgraded
Print throughput → Experiment reliability
Full capability → Accepted limits
Traded for reliability and usability
Theoretical metrics → Real prints
Testing & Results
Subsystems first, then the full print path.
Learnings
Material handling is the dominant constraint. Maintainability matters as much as print quality.
In Action
Printer in operation
Printed sample
Non-proprietary views only
Requirements don't come pre-written.
The product skill wasn't engineering. It was turning a vague research need into a buildable spec, then catching where the build diverged.