Semiconductor Printing: The Challenge of Manual Calibration
Scrona is a semiconductor printing startup delivering nano-scale precision for industrial electronics. Their custom printheads can handle any design, but one thing blocked scaling: calibration.
Each new project required manual adjustments of many parameters - a slow, expensive, and uncertain process. One experiment could cost CHF 30,000 and a week of downtime with no guarantee of success.
Relying on manual approach slowed down projects and put product quality and customer trust at risk.
Turning Human Judgment Into Measurable Data
Datali worked with Scrona in an AI & Data Strategy Workshop, using interviews, lab visits, and process mapping to determine if calibration could be automated and how to achieve it.
We found that Scrona’s experiments lacked proper data capture. Years of manual adjustments created valuable knowledge, but without documentation engineers had to start from scratch each time.
Building Reliable Processes That Scale
We proposed a roadmap with three pillars: building a data pipeline to log every experiment, using computer vision to evaluate print quality consistently, and applying adaptive experimentation to reach good calibration faster with fewer trials.
This strategy turns calibration into a repeatable, data-driven process. With this roadmap, costs drop, calibration speeds up, and engineers’ time is freed, allowing the enterprise to scale faster.
