Abstract: Industrial material movement in small and medium scale industries in India commonly depends on either low-cost Automated Guided Vehicles (AGVs) that follow fixed paths, or expensive Autonomous Mobile Robots (AMRs) with unnecessary sensing and processing overhead. AGVs lack path adaptability and halt when path conditions change, while AMRs are prohibitively costly for widespread deployment. This paper presents BHEEMA, a semi-autonomous navigation robot positioned between AGVs and AMRs. BHEEMA uses a modified region-focused Dijkstra algorithm running on a low-cost ESP32, along with a 5-sensor IR junction detection array for local navigation and Wi-Fi based coordination through a local server. The robot intelligently selects a subsection of the full map and performs local shortest-path planning, reducing computation while allowing dynamic routing. The system supports multi-robot coordination, collision prediction, and deadlock reduction inspired by the Braess Paradox. The design is highly cost-effective, scalable and suitable for Indian warehouse and cottage industry environments.

Keywords: AGV, AMR, Path Planning, ESP32, Dijkstra Algorithm, Swarm Robotics, Industrial Automation

Downloads: | DOI: 10.17148/IARJSET.2025.121237