At Agility3, the development of high quality, virtual 3D environments is one of our core capabilities. Typical examples of our work include:
- driving simulator environments to enable complex road layouts to be trialled during the design phase,
- interactive 3D Visualisations to enable stakeholders to experience the design of a new building and compare options, or
- interactive visualisations to show how safe access will be maintained during construction.
Figure 1 shows an example environment produced for a driving simulator trial. Further examples of our work can be seen on our portfolio page.
Figure 1 – Agility3 Driving Simulator Urban Virtual Environment Example
One of the crucial, early conversations we have with clients, when discussing a project, is around the data they have. The availability of good quality data describing the subject matter has a huge impact on the effort required to produce a 3D virtual environment as well as the quality and accuracy that can be achieved. With the emergence of BIM as a methodology of choice, over the last few years, we are often asked, “How can BIM data be used to create virtual environments?”.
BIM, or Building Information Modelling, is more an approach to collaborative working than a set of defined data standards or formats, but a BIM dataset will typically include 2D and 3D CAD models describing all aspects of the planned construction project that can be accessed by all stakeholders on a project.
The benefits of BIM Data reuse will vary enormously from project to project, depending on the data that is available, the quality of the data and the detail contained within it.
At a basic level, 2D CAD drawings (see Figure 2) can provide an excellent basis from which to develop a 3D virtual environment. Data provided must be to scale (ideally 1:1) and have good resolution so that curves are as smooth as possible. It must also contain all elements that should be represented in the 3D virtual environment, ideally on separate, appropriately named layers and must reflect the state of the real environment that is to be represented. It is possible to make changes to bring supplied CAD data up to date but this adds to the overall effort required.
We can work with a range of data formats, however we have found that DWG, DXF and SKP formats of CAD data are some of the easiest to reuse.
Figure 2 – 2D line CAD elevation drawing example (source: www,autodesk.co.uk)
From the example data shown in Figure 2 we would be able to efficiently create a 3D virtual environment that accurately depicts the planned new road layout. This could be used in a driving simulator or 3D visualisation application to trial the new layout, to improve communication, engage stakeholders and obtain more valuable feedback earlier in the project. Generally, as a project progresses, rectification of issues becomes more and more complex and expensive, so being able to recognise these issues early, using an accurate 3D model, makes modifications easier and more cost effective.
As already described, the type of 2D CAD data depicted in Figure 2 would serve as a good basis for the development of a 3D virtual environment. However, if 3D CAD data is available, then this provides a far superior foundation from which a 3D virtual environment can be developed. With 2D CAD data we often have to make a ‘best guess’ with the third dimension which can result in inaccuracies in the developed 3D model. For many applications, these inaccuracies don’t pose a significant problem and the usability of the 3D virtual environment is not affected. However, where accuracy and detail is of primary importance, 3D CAD data is invaluable. In addition, the availability of 3D CAD data allows us to more rapidly understand the requirements of a project and ultimately save time and effort in the design and development of a virtual environment.
The provided 3D CAD models can be imported into 3D modelling tools, such as 3D Studio Max, used by Agility3 and can form the basis of the virtual environment. This helps to ensure the virtual environment accurately reflects the real environment and CAD data is far easier to work with than a purely textual description of the environment or line drawings. Starting from the CAD, we would then optimise the data, adding visual details as required and applying high quality textures and materials to the environment. Finally, the environment is configured specifically for the end use requirements, whether that is for a driving simulator or interactive fly-through. Our in-house software development team are often required to create interactive applications using the virtual environments, for example, to switch between different design options, times of day or phases through construction. Figure 3 shows a virtual environment developed for the rail domain being used within an interactive application developed by our software team.
Figure 3 – Agility3 Rail Infrastructure Virtual Environment Example
In the absence of CAD data, a LIDAR scan of the area (see Figure 4) is a good substitute when you need to represent an existing environment. LIDAR is particularly effective for capturing fine detail and the precise positions of objects. Potentially, the 3D models that we would create for the virtual environment could be incorporated back into the BIM dataset and made available for all project stakeholders, thus contributing towards BIM compliance. The use of LIDAR is already widespread and continuing to grow. Currently, LIDAR point cloud data is traced within the 3D modelling tools to create an accurate 3D virtual environment. This process can be very time consuming and often requires more effort than working from CAD data, however, new commercially available tools are being developed which can automate some of the conversion from LIDAR point cloud to polygonal geometry. As these tools mature, modelling virtual environments from LIDAR will become quicker and easier, reducing the difference in effort required compared to working from CAD data.
Figure 4 – LIDAR point cloud example (source: http://www.mto.gov.on.ca/english/publications/road-talk/road-talk-19-spring.shtml)
The ideal form of BIM data that can be supplied to generate a virtual environment as accurately and efficiently as possible, is a pre-existing, low-polygon, accurate 3D model of the environment. This can be further optimised, representative textures applied and then configured to operate smoothly within the required 3D visualisation.
Ultimately, there are many forms of BIM data that can be used to help develop virtual environments and the available data can vary considerably in terms of quality, content and accuracy. This will impact the accuracy of the delivered virtual environment as well as the time and cost to develop it. Therefore, we would recommend that developers of BIM data:
- Consider all the potential stakeholders that may wish to reuse the data,
- Produce accurate CAD data in standard formats,
- Configure data to be easily understood with objects intuitively named and separated into different layers or groups and keep data up-to-date.
Good quality data can be reused multiple times, increasing the value of the data and the return on the data investment. High quality virtual environments can provide huge value to infrastructure and construction projects and the availability and reuse of good quality data can result in substantial savings. Therefore, when considering requirements for a simulator environment or interactive 3D visualisation, first consider the data that can be provided.
This article was written by Louis Drakes, Business Coordinator at Agility3. If you have any questions or would like any further information, please do not hesitate to get in touch at firstname.lastname@example.org or call on (+44) 01438 488066.