Jean-Francis Balaguer's

Virtual Environment and WWW for Engineering
Jean-Francis Balaguer
Computer & Network Division
CH-1211 Geneva 23

1. Introduction

CERN is about to start building the LHC (Large Hadron Collider), which at its completion, in the year 2004, will be the world's largest machine for High Energy Physics. The LHC will require huge investments in effort and funding from all the CERN member states, and it is of capital importance that each single aspect of this huge project can be analyzed in every little detail, so that errors in the design of the premises and equipment can be determined as soon as possible. This will require the most efficient organization and technology at every stage of the design process.

In order to evaluate and promote the use of virtual environment technology as a tool to help design, build and maintain the LHC premises and equipment, a pilot project was started in January 94 by the Computer & Network and the Accelerator Technologies divisions, under the name VENUS (Virtual Environment Navigation in the Underground Sites). In particular, VENUS tries to provide tools and services for virtual prototyping and networked design integration. Both of these projects are based on the i3D system that combines into a single tool the 3D input and high-performance rendering capabilities of high-end VR systems with the data-fetchning capabilities of networked browsers. It allows the exploration of three-dimensional scenes described using VRML [3], where each 3D object can be annotated with a document of any kind of media documents that can be accessed on the World Wide Web [4]. Using a 3D device, the user can explore its three-dimensional data and request access to other documents. Three dimensional data is handled directly by i3D while the handling of other types of media is currently delegated to NCSA Mosaic or Netscape.

2. Virtual Prototyping for LHC

In any project of the scale of LHC, the design phase is probably the most delicate one, as this is when some critical choices are to be taken which could dramatically affect the final results, timing and costs. Unlike in commercial manufacturing, accelerator designers and engineers don't have a chance to improve their product on a "second generation". Each instrument is built once with very small possibilities of modifications. This implies that it must be perfect on the first and unique trial.

The ability to visualize the model in depth is essential to a good understanding of the inter-relationships between the parts. Interactive navigation is a powerful method for reviewing, confronting and refining designs, that allows specialists to focus on items of interest, and simplifies identification of geometry in any section of the model. Colors assigned to engineering disciplines help examine design integration, look for gaps, part interferences and mismatches. Since the premises and equipment can be electronically preassembled, all fixing interferences can be visually checked, even clearances necessary for installations or moving parts and, all this accomplished well before the final design is released for construction. Since changes made in the later phase of production cost many times more than those caught in the prior phases, savings are inevitable.

Virtual prototyping allows managers and engineers to critique design as when watching physical mock-ups, but better and earlier. Virtual prototypes can be made available easily and immediately, at no extra cost, since each part has already a three-dimensional computer representation inside the CAD database. Since it can easily be kept in sync with the current design, virtual prototyping gives us much more of a responsive ability to study how the various parts and assemblies fit together; it helps reduce risks and costs, and shorten schedule.

3. Networked Design Integration

The LHC-system will be designed, engineered and manufactured in a distributed manner by about three hundred institutions around the world. One of the challenges to be taken up, in order to guaranty the success of the LHC project, will be that of the technical information exchange for design and documentation purposes, between all the partners involved [2]. In particular, being able to access current or previous designs, in a seamless manner, is essential.

The World-Wide Web has added a universal organization to the data made available on the Internet allowing to view all hosts as a unique data source, and to treat all of this data as part of a single structured document. By composing the descriptions of each part or subsytem, and accessing them directly from the site where they are being designed, we are able to retrieve virtual prototypes reflecting the status of the latest or some previous version of the design. This behavior can be easily obtained by exploiting the inlining capabilities of both VRML and i3D's file format.

The browsing capabilities of the i3D system allows users to interactively recall and view information attached to 3D models, by selecting objects of interest during navigation. By allowing users to interactively recall and view the attached information by selecting objects of interest during navigation, the interactive 3D viewer becomes a natural front-end for querying information about 3D models. Annotations can refer to text, still images, technical drawings, sound, animations or even other 3D models, exploiting in this way all of the digital media capabilities of current workstations. Annotating each three-dimensional model with an HTML page provides us with the basic ability to structure the available technical information. Soon, we will be able to build such pages automatically when exporting the prototype from EUCLID. In the future, we plan to investigate the association of multiple URLs with each three-dimensional model and to provide context sensitive selection to determine which URL needs to be retrieved.

4. Conclusions and Future Work

In this paper, we have presented how, by combining into a single tool the 3D input and high-performance rendering capabilities of high-end VR systems with the data-fetching capabilities of network browsers, we are able to handle the distributed nature of the LHC project and address some of its information management issues. In particular, by accessing individual parts and subsystems directly from the site where they are being designed, we are able to reflect the distributed structure of the design process and make available virtual prototypes that are always in sync with the latest design. The browsing capabilities of the system allows us to attach documents of any
kind of media to each three-dimensional objects, offering a means to structure technical information and to present it under the most suited media.

Future work will concentrate on addressing the specific issues to the use of virtual environment technology in conjonction with the WWW for engineering purposes. In particular, we should consider the following directions of work:

  1. What are the functionalities that a browser have to offer in order to address the problems related to engineering? Do we need animation? Do we need geometric queries? Do we need specific visualization techniques?
  2. What is the best way to retrieve information? Do we need context sensitive hyperlink retrieval?
  3. What is the level of experience sharing that is required? Do we need multiple networked participants? Are settings for design meetings enough?


[1] Balaguer J-F, Gobbetti E (1995) i3D: a High Speed 3D Web Browser. Proceedings VRML'95, San Diego, California.

[2] Bachy G, Hameri A-P, Mottier M (1995) Engineering Data Management: a Tool for Technical Coordination. CERN MT/95-07, LHC Note 345.

[3] Bell G, Parisi A, Pesce M (1995) The Virtual Reality Modeling Language. Available through WWW at

[4] Berners-Lee TJ, Cailliau R, Groff JF, Pollermann B (1992) World-Wide Web: The Information Universe. In Electronic Networking: Research, Applications, and Policy 2(1): 52-58.

[5] Brooks FP Jr, Frederick P (1986) Walkthrough - A Dynamic Graphics Systems for Simulating Buildings. Proc. SIGGRAPH Workshop on Interactive 3D Graphics: 9-22.

[6] Frystyk H, Lie WH (1994) Towards a Uniform Library of Common Code. Proc. WWW Conference, Chicago, USA.

[8] de Gennaro S (1995) Virtual Prototyping at CERN. Proc. CHEP'95, Rio de Janeiro, Bresil.