Back to all work

Research prototype / spatial computing

Industrial AR Assembly Guidance

AR guidance for industrial assembly using HoloLens 2, Apple Vision Pro, object detection, spatial overlays, and structured instructions.

Apple Vision ProPublished HCI ResearchCore ML Computer Vision
  • Vision Pro
  • HoloLens 2
  • Core ML
  • ARKit
  • LLMs
  • SPARQL
Exploded 3D compressor model with contour guidance overlays
Domain Industrial HCI
Devices HoloLens + Vision Pro
AI layer Core ML + LLMs

What the project needed.

The work focused on helping operators follow complex assembly procedures with spatial guidance instead of static manuals.

The prototypes combined camera input, detected parts, visual overlays, structured domain knowledge, and voice-driven queries.

What I shipped.

  • Built AR assembly prototypes for HoloLens 2 and Apple Vision Pro.
  • Used Core ML object detection and contour overlays to guide users through physical assembly steps.
  • Connected SPARQL knowledge graphs and LLMs so operators could query instructions in natural voice.
  • Published academic research on spatial computing in industrial human-computer interaction.

Technical architecture & leadership.

  • Designed the spatial tracking and Core ML visual alignment algorithms.
  • Built natural language query prototype interfacing with RDF/SPARQL ontology databases.

Results in production.

  • Published in leading HCI venues; highly praised by research partners for rock-solid spatial stability.

Have a similar challenge?

Whether you are architecting a new Apple product, optimizing video/AVFoundation pipelines, or enforcing strict TDD, tell me what needs to be solved.

Discuss your project