Securade.ai HUB - A generative AI based edge platform for computer vision that connects to existing CCTV cameras and makes them smart.

AI-powered computer vision system for real-time workplace safety monitoring. Detects people and PPE compliance (helmets, vests) using YOLO models with intelligent tracking and MP4 output.

AI-powered safety monitoring system for industrial sites, detecting PPE compliance and unauthorized access in real-time with facial and object recognition.

An AI-powered CCTV surveillance system for real-time detection of PPE compliance, including helmet and mask violations, using YOLO and computer vision.

Local AI automation workbench for Siemens PLC and industrial-robot R&D: MCP, TIA Portal V21, PLCSIM, guarded writes, and offline-first verification.

嵌赛地瓜机器人RDK X5赛道仓库：基于RDKX5的自训练模型可视化主动安全干预系统与双模低载LLM推理

Problem Statement : Aerial Location of Hazardous Atmosphere in Industries (Ministry of Home Affairs).

Desenvolvimento de um Sistema de detecção de EPIs (PPE) utilizando YOLO para identificação automática de equipamentos de segurança em imagens.

An advanced real-time object detection and tracking system for industrial safety.

This is a computer vision project that uses YOLOv9-e model and checks if workers wear their Personal Protective Equipment (PPE) and if they keep a safe distance from work machines.

NLP based Industrial Accident Severity Assessment

Desktop-first Unity safety trainer demo for oil production site inspections with VR-ready architecture.

An AI-powered CCTV surveillance system for real-time detection of PPE compliance, including helmet and mask violations, using YOLO and computer vision.

Bare-metal security research, packet-level dissection, and detection engineering across OT/ICS industrial communication protocols. No high-level libraries.

Web app para monitoreo de seguridad industrial en tiempo real: detecta el uso de cascos (PPE) en imágenes y video usando YOLOv8 y FastAPI, con alertas automáticas por WhatsApp.

AI-powered industrial safety platform that detects mismatches between shift-handover language and live sensor data — preventing accidents in the critical post-shift-change window.

Real-time compound risk detection for industrial plants — fuses gas sensors, work permits & worker location into one live risk score. Built for ET AI Hackathon 2026.

Fully-local RAG assistant for Indian process-safety documents (OISD, PNGRB, PESO, NDMA, DISH): grounded, cited answers, hypothetical training drills, and pre-handover permit checks — every claim traceable to source.

ML surrogate models (ANN, Random Forest, MLR) that replace ALOHA dispersion simulations for real-time hazard-zone radius prediction — 1,215 full-factorial scenarios, with out-of-distribution robustness analysis.

Discussion paper exploring passive visual guidance systems (photoluminescent materials, reflective delineation) as a supplementary safety layer in offshore, marine, energy, and industrial facilities - covering escape routes, muster stations, helidecks, and shipyard/dry dock operations under degraded visibility conditions.