Collaborative design review in a distributed environment
Abstract:
This paper investigates how collaborative environment technology can be used to integrate geographically dispersed design teams and other technical and non-technical members from within and outside the organisation to support collaborative product design reviews. The creation of such a collaborative environment for design will allow participants from different phases of the product development cycle to collaborate and form a virtual team. A new framework is proposed to extend the current advances in collaborative technologies to provide access to local and global resources within design teams and the deployment of private and public design review workspaces to provide an efficient collaborative design environment. Also, a prototype collaborative CAD environment has been developed and the results are discussed in this paper.
Keywords: Collaborative Design, Shared Workspace, Design Review
By: Manglesh Sharma (a), Vinesh Raja (a), Terrence Fernando (b)
(a): University of Warwick, Warwick Manufacturing Group, UK, {manglesh.sharma,vinesh.raja}@warwick.ac.uk
(b): bUniversity of Salford, Future Workspaces Research Centre, UK, t.fernando@salford.ac.uk
login or register to download the paper.
Thanks for asking those questions.
1)It is for conducting design reviews at first place and accommodating less time consuming modifications at the time of review itself to reduce the design cycle time. It allows the collaboration among design team members. It allows two kinds of shared workspace: private and public utilising two kinds of resources: local and global. These terms are explained in the paper.
The proposed system is real-time system and uses multi systems like CAD/Assembly/FE analysis (interface based access) during collaborative design review. The owner of the cad system shares the models with other participants. The owner will be able to edit the model in the CAD system itself and the changes will be reflected immediately because the plug-in is in direct connection with the CAD system.
Secondary advantage, the proposed system also allows using not only the CAD system but also other kind of systems like FE/Motion/Assembly packages. In case of global resource sharing (central cad system) the model loading/unloading can be done remotely. Besides just models sharing it permits display of replication of the features based tree structure as it can been in the CAD system. A CAD plug-in this respect was built which allows connective live to SolidWorks CAD System and sharing the data with others and editing the model (by owner of CAD system, also termed as local resource user) and reflecting the changes to other participants. Plug-in for other kind of packages will allow broader collaboration. On communication side, different types of modern tools can be used.
2) The virtual reality model (tessellated) is used only for the means of display. Any editing operations over the network are time and resource consuming. The real owner will be the ideal in this practice as he has the system as local to him and editing will not take longer and changes will be displayed back to individual because of the live connection using the plug-in. Secondary advantage is that the owner has the legal mathematical model as this is not being distributed and this will be important if the participants are outsiders to the organisations (consultants??) and needs only his valuable feedback to expedite the process. Plugins for different type of CAD systems or other packages such as FE/Motion/Assembly analysis can really help to create such an environment to bring this wide market under one roof.
I hope this answer will suffice the expectations.
Thank you.
Dear Manglesh and all,
Many thanks for your reply. I have another question related to the graphic displaying speed in team member's PC when there are changes in CAD models of the owner's PC via plug-in.
Since collaborative design projects normally require lots of works, and the complexity of the project is big. In the product development, CAD data of the product are normally in assembly or complex with a big file size. We experienced product projects in which STEP or IGES format data are normally about 10 to 20 MB, some may be about 30 to 40 MB.
The complex and big file size of the CAD models lead to reducing the displaying speed in the computers where the team members who are geographically distant and working with the owner of the CAD data. This is because it takes time to convert the CAD models to the triangle mesh ones before it can be displaying.
The question is as follows:
”did you consider or test the displaying speed of the proposed system with different CAD file size and complexity?”
It would be good if this information is included in the paper.
Thank you again,
Le, Cardiff University, Wales, UK.
I had worked on developing the plug-in to connect it to SolidWorks. It worked in displaying the model on a remote pc (P IV, 512MB RAM) over intranet during testing. Regarding speed, yes we had found it was slow, but the actual reason could not be ascertained because of the inhouse development of framework (Hydra framework- University of Salford) was not complete and still under development of some module.
Also the plug-in module was written in VB (for SolidWoks API) which is slower compared to C++. So when we tested on a 5MB CAD model data it took around a minute to display. The model shown in the Figure 4 in the paper took 7 seconds to appear on the screen.
The improvement could have been achieved by porting the code to c++ and improving the scenegraph rendering module of HYDRA. At the end there was one more suggestion made to use the HOOPS 3D Application Framework (HOOPS/3dAF) instead which looked promising. But because of shortage of time no more work could be done to achieve the much better speed as desired.
But definitely, technology (like different application framework/software/hardware, some more experiments) can use the proposed thought explained in the paper to overcome this deficiency sooner or later in the near future.
Hi Manglesh,
Many thanks for your clear answers to the mentioned questions.
I think that the success of the collaborative design software or plug-in (concretely your proposed system) would depend mainly on the following factors or issues:
1- The displaying speed: as mentioned by you, this may be solved technically by moving to stronger Computer Graphic programming languages. I think the algorithms also need to be taken into account, in most cases, it is the main issue for the programmers.
2- The interactive capability of the design team members (clients) to the owner’s CAD system: It would be good if the plug-in (to Solidworks) can generate a collaborative environment for the design team members (clients) in which all the members can use the same CAD packages (here is Solidworks) to directly communicate with the owner’s CAD system, such as the one done by UG NX or Truespace software.
3- Since you mentioned to the future development of the plug-in or the same system that are capable of handling different types of CAD/CAE/CAM data, the data interface problem would be a main issue to be taken into account. Moreover, in these systems, not only the graphical data are shared, but also the analysis information as well as manufacturing features are included.
Looking forwards to seeing your comming research results!
Best wishes,
Le, Cardiff University, Wales, UK
Dear Manglesh and all,
I have two questions and comments related to “collaborative design” .
Question No.1 : Most of the collaborative CAD and 3D modeling packages such as UG NX and TrueSpace (from version 7) allow the members in the conferences (or meetings) view, manipulate and edit 3D objects. Your proposed system that was implemented in Solidworks package uses the Virtual Reality models (*.VRML files) for display purpose; and a native Solidworks CAD format (*.asm files) is not utilized. It means that the conversation from a native Solidworks CAD models to triangle mesh models (VRML format) is needed; and members in the conferences (meetings) can not edit 3D models. Therefore, my first question is
"In views of “collaborative design” , what are the strong points of the proposed system compared to the commercial packages such as UG NX and TrueSpace?"
Question No.2 :
3D Triangle Mesh models such as STL and VRML are commonly used for product review by geographically distant clients (customers); the typical commercial package using triangle mesh format for this purpose is Materialise NV’s products (Magics Communicator). Magics Communicator allows online conferences and discussions about the design in real time via Internet, LAN or modem. The input data format can be STL, MGX, DXF 3d faces*, SolidEdge*, IGES*, VDA*, ProE*, CATIA, Unigraphics and native Parasolid.
Recently rendering 3D models are able to be viewed and displayed online using internet browsers; this is used for design review as well as marketing purposes. A few examples of viewing the rendering 3D products is at this WEBSITE (click HERE to view online viewing of rendering 3D design)
It is very difficult for customers (clients or team members) collaboratively participate in 3D modeling and modification using triangle mesh formats because of the limitation in 3D data manipulations. In addition, it takes time and needs Reverse Engineering processes for the conversion from triangle mesh models to CAD data, and this leads to the loss of designer’s intend; and when the RE processes are used, it is not a collaborative design environment anymore. This is why “single” CAD packages such as UG NX and Pro/Engineer are developed and used for a collaborative design purpose. Therefore, my second question is
"Do you think that the development of the plug-in modules to available CAD packages is useful for collaborative design? If YES, what is the solution to create a collaborative design environment? "
Thank you,
Le C.H, Cardiff University, Wales, UK.