Ian Spehar

Ian Spehar

MS in Mechanical Engineering and Robotics at OSU

SCAPE: Automated Process Control for Laser DED Additive Manufacturing

title: “SCAPE: Synchronized Process Control for Laser Directed Energy Deposition” excerpt: “Software framework for automating synchronized robot motion and real-time process parameter changes during laser DED additive manufacturing” order: 1 collection: portfolio image: /images/scape_setup.jpg —

SCAPE (Scheduled Control for Automated Parameter Execution) is a software framework I developed for my M.S. thesis that automates synchronized process control for a Kuka-Meltio Laser Directed Energy Deposition (LDED) system. The framework coordinates 6-DOF robot motion with real-time laser power and material feed rate changes at precise physical locations, enabling generation of labeled manufacturing datasets for machine learning research.

Key Contributions:

  • Designed and implemented control software architecture in Python
  • Integrated with KUKA robotic manufacturing platform and Meltio printer
  • Enabled systematic generation of labeled datasets for ML research
  • Validated through additive manufacturing trials and statistical analysis
  • Achieved reliable synchronization with measurable geometric changes

Technologies:

Python · KUKA Robotics · Meltio 3D Printer · LDED · Motion Control · Manufacturing Automation · Experimental Design · Statistical Analysis

System Architecture:

Two-stream coordination: SCAPE synchronizes (1) 6-DOF KUKA robot motion with (2) laser power and material feed rate changes, triggering parameter updates based on the robot’s physical position during printing.

Data generation: Logs synchronized position, laser power, and feed rate data at each physical location, enabling labeled dataset creation for process monitoring and ML applications.

Experimental Validation:

  • Designed trials with controlled parameter variations (laser power, feed rate)
  • Collected synchronized process data and post-print geometry measurements
  • Performed statistical analysis confirming significant geometric changes between process states

Results:

✅ Reliable synchronization across multiple print trials ✅ Measurable geometric changes corresponding to commanded states ✅ Systematic dataset generation for ML research ✅ Statistically significant process-structure relationships

Lessons Learned:

  • Real-time coordination requires careful timing and communication management
  • Industrial integration demands understanding of proprietary protocols
  • ML-ready datasets require thoughtful experimental design

[Add link]

Code:

[Add link if available]