Vision-Based Multi-Camera Robotics Simulation: The SIMLAN Open-Source Framework for Industrial Environments

Authors: Hamid Ebadi, Pär Aronsson

Affiliation: Infotiv research, Gothenburg, Sweden

Venue: Submitted to the 19th PErvasive Technologies Related to Assistive Environments (PETRA) Conference (Under Review)

Abstract

We present SIMLAN (SIMulation for Indoor Multi-Camera Localization And Navigation), an open-source simulation environment designed for rapid prototyping and testing of multi-camera robotic algorithms.

Inspired by Volvo Group’s Generic Photo-based Sensor System (GPSS), the SIMLAN project models ceiling-mounted camera networks and factory layouts to enable global, vision-based monitoring, localization, and navigation. The system includes utilities for:

Deep-learning-based visual human pose capture and replay
Bird’s-eye image stitching
Geofencing
ArUco-based localization and navigation

All packaged in reproducible Docker containers. SIMLAN reduces the barriers to prototyping and research on the monitoring, verification, and validation of vision-based robotics systems.

Keywords: Simulation environments · Robotic · Computer vision · Multi-camera systems · Human–robot interaction · Verification and validation

Citation

Ebadi, H., and Aronsson, P. (2026). Vision-Based Multi-Camera Robotics Simulation: The SIMLAN Open-Source Framework for Industrial Environments. Submitted to the 19th PETRA Conference. (Under Review)

Download PDF Download BibTeX GitHub Repository