Abstract—People with disabilities often face critical challenges in accessing timely and comprehensive emergency alerts. This study presents an adaptive multimedia system that uses safety engineering principles using Web Content Accessibility Guidelines (WCAG). Based on WCAG, the system was created to meet visual, auditory, and tactile output requirements based on user profiles and environmental conditions. The system inputs are fed through the Internet of Things (IoT) and processed by a rule-based adaptation engine. The emergency adaptive system combines environmental data, user health, and manual input to generate appropriate multimedia alerts (audio, visual, or vibration). A prototype was built using React/Tailwind for the user interface and integrated with sensors for environmental monitoring. The system was evaluated in an emergency scenario on 32 participants including groups of people with visual, hearing, cognitive, and multiple disabilities. The evaluation results showed a high average system usability scale (SUS) score of 86.2, indicating excellent usability. The average Accessibility Coverage Metric (ACM) score was also obtained at 0.89, indicating broad and inclusive modality support. Based on the experiments, the Adaptive Decision Function (ADF) score exceeded 0.90 for combinations such as visual with vibration and audio output. This shows strong adaptive performance under various conditions. These results demonstrate the potential of the system to improve the inclusiveness of emergency response for people with disabilities.