A city’s traffic cameras, drones, and dash cams will feed a single virtual-reality control room that lets authorities see and manage entire urban environments in real time. This matters because cities currently rely on fragmented systems that cannot share data during crises. When a flood or terrorist attack knocks out local networks, traffic management and emergency response often fail in the first critical hours. IMMERSE solves that by building a unified digital twin of a city—a live, 3D replica updated by computer vision that transmits only relevant information, not full video streams, preserving privacy through federated learning. Satellite links keep the system running even when ground networks go down. If successful, the platform could let a single control centre reroute traffic around a collapsed bridge, guide delivery drones away from a fire, and track pedestrian flows to prevent stampedes—all within minutes. The project builds on an existing prototype, so it aims for rapid, scalable deployment in UK and Singaporean cities rather than years of fundamental development.
View original technical description
IMMERSE expands upon the successful Innovate UK project Epi-terrestrial Multi-modal Input Assimilation (EMIA-10036013) to deliver a transformative system for urban mobility, traffic optimization, and disaster management. By integrating immersive digital twins, computer vision, federated learning, and robust network technologies, IMMERSE creates a unified framework to address urban challenges while safeguarding privacy and enabling super-intelligent AI systems. The system utilizes data from drones, custom dash cameras, and CCTV systems to recreate dynamic urban environments and scenes in real-time within an immersive VR interface. Computer vision supports this task by detecting and transmitting only the relevant information and not the whole video streams. This information is used by the digital twin to recreate the environment and provide actionable insights for disaster management, traffic forecasting, and pedestrian flow, enabling city authorities to make informed decisions. The modules from the proposed system, including weather tracking, traffic analysis, pedestrian movement monitoring, disaster detection, drone collision avoidance, scene analysis and optimal drone routing, work collaboratively within the platform to enhance operational accuracy (e.g. scene analysis supports/is supported by pedestrian movement, weather detection supports traffic forecasting etc.). The use of federated learning ensures the computer vision modules continually improve through decentralized data exploitation without compromising individual privacy while exploiting incoming data from multiple sources (every drone, CCTV and dash camera will contribute to this). IMMERSE also introduces advanced network technologies, such as satellite links, to maintain operations during critical events, including network outages. By supporting traffic management and drone mobility for deliveries and public transportation, the platform facilitates smart city mobility and traffic flows even during disasters, ensuring resilience in the first crucial hours of crisis management. The project builds on EMIA's existing VR environment, custom dash-cameras, developed backend structure, traffic forecasting and pedestrian movement detection, and proven modular architecture. For this reason, IMMERSE minimizes development time and costs while providing a scalable, cost-effective solution for cities worldwide. The project benefits from partnerships with organizations like Singapore's SLA and LTA (Singapore\_Land and Transport\_Authorities) and UK city councils, combining extensive market research with proven smart city technologies to establish a new standard for urban mobility and disaster resilience, and the expertise of Interactive Coventry (IC) in developing AI based solutions and delivering large scale smart city projects. By integrating privacy-preserving AI and immersive tools, IMMERSE aligns with global sustainability goals, enhancing urban safety, mobility, and disaster preparedness while respecting privacy and creating a framework for next-generation intelligent systems.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know