RapidScan, snelle en autonome AI-generatie van 3D-omgevingen
RapidScan is a tool for quickly creating realistic 3D simulations of real-world locations, generated from photos or videos of the site. Once a location has been captured, the resulting scenes can be viewed and used in a VR headset within minutes. RapidScan builds on earlier work by Leiden University with 3D Gaussian Splatting, which demonstrated that complex urban areas can be reproduced in remarkable detail. By focusing on speed and ease of use, the tool offers a practical way to explore or rehearse situations, opening up opportunities for the Netherlands Police, municipalities and others with an interest in simulating physical environments.
Involved organisations
More projects from Innovation Impact Challenges
Discover other projects within this funding scheme
Planning & Real-time IX Scenario Maker
PRISM is an AI-supported VR scenario builder that lets police officers with no technical background build and adapt realistic environments and scenarios. Using voice commands and simple building blocks, they can recreate past incidents and prepare for new situations, for example crowd control or threat scenarios. The tool drastically cuts the development time for XR training and makes it possible to share and reuse scenarios nationwide. In phase 1, the team is investigating the technical feasibility, user acceptance and economic viability of a proof of concept for this AI-driven IX scenario builder.
Scenebrook Haalbaarheidsonderzoek Interactieve XR-Trainingsscenario’s
The project explores whether Scenebrook can be extended with features that let the Netherlands Police develop VR training scenarios faster and more accurately. We're testing three things: automatically converting geo-data into realistic 3D environments, building complex, police-specific scenarios without requiring technical expertise, and measuring learning outcomes in VR. The aim is to determine whether this would allow the Netherlands Police to create realistic, up-to-date and easily adaptable training for situations that are difficult to practise physically.
Training Sandbox
The Training Sandbox is a hybrid IX platform that lets operational police officers independently create, adapt and share realistic scenarios within hours. A web editor makes it possible to quickly recreate ad-hoc situations and large-scale events, while a Unity package gives in-house development teams room to customise and expand the platform. It supports real-time intervention, learning analytics and multi-user collaboration within a single simulation. Built on a privacy-conscious architecture and guided by public values, the Sandbox aims to strengthen officers' practical skills, professional competence and mental resilience, cutting turnaround time from weeks to hours.
AI-Avatar
Police officers are constantly confronted with unexpected situations that require them to act quickly, provide assistance and gather information. Realistic training is essential for this, but it's expensive, time-consuming and hard to scale. With an AI avatar – a digital, human-like figure driven by artificial intelligence – officers can practise realistic scenarios using VR headsets. The avatar switches flexibly between roles such as suspect, witness, victim or colleague, responding naturally to the officer. This allows officers to train their communication, judgement and decision-making in a safe, true-to-life environment, making it easier, more efficient and more scalable to practise for unpredictable situations.
Situational Awareness op elk moment, door iedereen
With Imagem NL B.V.'s innovation, users without specialist knowledge are able to independently create and explore a digital, photorealistic copy of a location. The location is based on up-to-date data such as drone footage, laser scans, 360° video, live feeds and administrative sources. Within this 3D virtual world, users can analyse, simulate and coordinate situations — think of line-of-fire and line-of-sight analysis, or working through 'what if' scenarios. This new information position enables them to create a realistic digital environment for preparation, coordination and action in both ad-hoc and planned situations. The result: faster decision-making, greater situational awareness, more efficient use of resources, and a safer working environment for officers on the street.
Multifunctioneel AR-systeem voor Real-time Visuele herkenning
The system uses AR glasses with built-in cameras and powerful AI to automatically recognise number plates, vehicles and personal characteristics from the officer's field of vision. This makes the system particularly well suited for use within the TBB, TBP and forensic investigation domains. Officers can also use the system for operational tasks on the street, where real-time recognition adds crucial operational value. Information is projected directly into the glasses, allowing officers to work hands-free. The system is controlled through subtle hand and eye movements or app integrations, enabling rapid recognition and decision-making during surveillance, incidents and traffic checks — all while maintaining continuous situational awareness.