Technology

The success of our methodology lies in our innovative modeling processes, which are draw from our experience of the gaming industry. They enable high levels of efficiency and make our 3D models extremely data-light and performant.

NOTE! This site uses cookies and similar technologies.

If you not change browser settings, you agree to it. Learn more

I understand

For each 3D model, a database is required, but data capture processes can be very different in terms of effort and cost. We have developed a technology based on “terrestrial photogrammetry” to create high quality, accurate models, which were previously very expensive to produce. Ordinary images taken with a digital camera in combination with a number of reference measures are all we need. Here are two examples: it takes around three man-hours to undertake a full data capture exercise of all public areas, including outdoor areas, station concourses, all platforms and pedestrian tunnels, at a city center station with around 60,000 passengers per day and 14 long-distance tracks. The data capture process for a city with a population of 160,000 takes around 16 man-hours. Of course, we can also process data from other sources and technologies, such as laser scans, point clouds, traditional surveys, CAD, GIS, etc. - from quick-but-not-at-all-dirty to high-precision as required.

3D modeling is a semi-automated process based on our ToolChain, a compilation of algorithms for processing the input data. First of all vectors and then objects are identified in the digital model and are replaced by corresponding items from the object library – this ensures that the dataset remains small. We put in place, amongst other things, data synchronization processes and standardization, terrestrial and photogrammetric calculations (Matrix calculations) and detection software (pattern recognition) as well as library architecture and instancing.

With our multi-stage process, we achieve a resolution of ten centimeters, which can be improved to millimeters, depending on the quality of the input data. Individual buildings or structures are modeled manually by our experienced team of over 90 trained modelers.

Our models can either show the existing natural and built environment, the design options, or a combination the two.

At the heart of our technology is our comprehensive and well-structured object library consisting of geometric representations of the built and natural environments. It has been developed over many years and from many thousands of projects and includes up to eleven degrees of detail to deliver different specifications in the best way.

The library is organized on three levels. On the first, local level are individual objects, single houses for example. The second, regional level includes country- or domain specific objects such as traffic signs, railway signals or street furniture. The third, global level contains items that can be found anywhere in the world, for example, materials, surfaces, trees, cars, etc. Special items, such as ornamental facades are not included, but are separately photographed and manually modelled. Other identifying features are removed by our system so that there are no data protection or legal issues.

The structure of our object library allows our models to connect to the facilities or asset management systems of owners and operators through a simple mapping method. Comprehensive BIM models are produced from the object-based 3D models, independent of the software being used.

Many of our customers want to be able to view and operate our 3D models without having to install special software. We, therefore, supply the models for visualization and communication in our own "player", the 3D Rocket data viewer. However, we can also provide the models in all common software formats (CAD, BIM, GIS), so that planners and engineers can edit the data.

Our technology is as innovative as it is flexible and enables us to focus on our client’s needs and requirements