Aliaksei Petsiuk

Additive Manufacturing R&D

Projects


1. OrcaSlicer — Non-Planar Slicing & Tool-Change Optimization
Description: Fork of OrcaSlicer where I develop non-planar slicing and multi-color tool-change minimization for FFF printing, extending graph-based toolpath optimization to next-generation slicer architecture.
Link: https://github.com/apetsiuk/OrcaSlicer

2. PrusaSlicer — Interlayer Tool Clustering
Description: Original implementation of an interlayer tool-clustering algorithm for multi-color tool-change minimization — the first working version of the graph-based approach later extended in OrcaSlicer.
Link: https://github.com/apetsiuk/PrusaSlicer

3. FullControl G-Code Patch

Description: Computer-vision-based G-code patching: detecting a damaged region mid-print and generating a corrected toolpath to repair it in the following layer.

Link: https://github.com/apetsiuk/FullControl_GCode_Patch

4. Title: OctoPrint Failure Analysis

Description: An OctoPrint plugin for monocular computer vision-based, layer-wise defect detection in FFF 3D printing — in-process failure analysis during a live print. (In progress.)

Link: https://github.com/apetsiuk/OctoPrint-Failureanalysis
 
5. Title: Microstructure Vision-Based Porosity Analysis

Description: A computer vision pipeline for detecting and quantifying porosity in aluminum alloy microstructures from microscope scans, extending defect-detection methods to post-process material characterization.

Link: https://github.com/apetsiuk/microstructure-vision-based-porosity-analysis

6. Title: G-Code Parser and Visualization

Description: A Blender-based G-code parser and photorealistic renderer that generates labeled toolpath visualizations with segmentation masks, used to build synthetic training data for computer vision models.

Link: https://github.com/apetsiuk/GCode-Parser-and-Viz
 
7. Title: Raccoon — Camera and LIDAR Robotics Quadrathlon

Description: A multi-sensor (camera and LIDAR) navigation and perception system built on a Clearpath Jackal robot in ROS — foundational work in sensing and control for robotic manufacturing systems.

Link: https://github.com/apetsiuk/Raccoon

Robot with vision-based control


Development of a ROS package for remote control of a robot equipped with a computer vision system.


Arduino motor control


Description


4F optical correlator simulation


The system is based on the Fourier transforming properties of lenses. These properties give us an opportunity to implement direct and inverse Fourier transforms of two-dimensional objects and modify and analyze their spatial frequency spectra.