Multi-camera spatial reasoning for changing industrial environments
Industrial environments are dynamic: equipment and personnel move, occlusions are common, and a fixed geometric rule can quickly stop reflecting the state of the floor.
Our system combines four camera perspectives with a rule-based spatial reasoning engine that updates safety-zone boundaries from detected blocker positions. Rather than treating each camera frame independently, the system turns distributed visual observations into a continuously updated representation used for hazard monitoring.
My contribution was designing and implementing the spatial reasoning logic that determines how detected blocker positions redefine the active safety boundary. The detection pipeline was developed by a co-author.
This work shaped my interest in agents that must maintain task-relevant understanding of changing environments before they can make useful decisions.
We present a multi-camera perception system for dynamic safety zone reconfiguration in active industrial environments. Static safety zone definitions fail in dynamic industrial floors where equipment and personnel positions shift continuously. Our system fuses four simultaneous camera perspectives using a rule-based spatial reasoning engine to redefine green zone boundaries in real time based on outermost detected blocker positions, enabling proactive hazard detection before incidents occur.
@inproceedings{toth2026safety,title={Development of Trialing Image Detection for a Melt Shop Safety Tool},author={Toth, Kyle and Polra, Jay and Chauhan, Dhwanil and Ni, Yang and Zhou, Chenn and Fisher, Conrad and Page, Garrett},booktitle={AISTech 2026 — Iron and Steel Technology Conference Proceedings},year={2026},url={https://dhwanil832.github.io/publications/},note={Purdue University Northwest · CIVS · SMSVC}}