A System Architecture and Simulation Environment for Building Information Modeling In

A System Architecture and Simulation Environment for Building Information Modeling In

A System Architecture and Simulation Environment for Building Information Modeling in Virtual Worlds

John Oerter

Wyatt Suddarth
Matthew Morhardt
James Gehringer / Michael L. McGinnis
Johnnie Shockley

Army Corps of Engineers
Peter Kiewit Institute,
University of Nebraska
1110 S. 67th Street
Omaha, NE 68182-0694 / Allison Baysa

SAIC
6825 Pine St., # B10
Omaha, NE 68106

Keywords: Building Information Modeling, BIM, Virtual Worlds, Virtual Environments

Abstract

This paper describes building information modeling (BIM)research by students and faculty of the University of Nebraska (NU) Peter Kiewit Institute (PKI) in partnership with Science Applications International Corporation (SAIC). The objective of this on-going research is to create a virtual modeling environment where architects and architectural engineers can present building concepts and design options to customers in a way that is more easily envisioned by the client. The software used for this research project is Autodesk® Revit, Autodesk® 3DS Max,SAIC’s 3-dimensional(3D) virtual world calledOnline Interactive Virtual Environment (OLIVE), versions 2.4.0 and 3.0, anda proprietary game engine called Unity developed by Unity Technologies based in San Francisco, California. This paper overviews development of a system architecture in which BIM will be integrated with a virtual simulation environment built in OLIVE and research results to date.

  1. BACKGROUND AND RELATED RESEARCH

PKI occupies a 192,000 square foot building on the University of Nebraska at Omaha (UNO) south campus. The Institute is home to approximately 85 faculty and 11 academic programs serving 1,800 students from two colleges: University of Nebraska-Lincoln (UNL) College of Engineering and the UNO College of Information Science and Technology. Other units operating out of the Institute are the Holland High Performance Computing Center and the Peter Kiewit Institute Technology Development

Corporation. Schools and research centers based at PKI include the Charles W. Durham School of Architectural Engineering and Construction, the School for Interdisciplinary Informatics and the Nebraska University Center for Information Assurance among others.

A literature review identified various papers on virtual environments used to enhance building information modeling (see [1], [6], [7] and [9]). The literature review also revealed other applications not directly related to BIM but which may have relevance to architects, engineers and educators working in areas such as building sustainability and energy harvesting, first responder training, and interactive online education. Alahmad, Nader, Cho, Shi and Neal [1], for example, presented research and findings on building information models developed to monitor energy consumption. Ruppel and Schatz [7] discussed integrating BIM and virtual environments to create a realistic training simulation where firefighters can practice how to clear buildings or train on firefighting techniques while navigating through virtual buildings. Ku and Mahabaleshwarkar [6] discussed merging BIM and gaming worlds which the research team found particularly helpful to their collective understanding of how to couple BIM and OLIVE. Finally, Shen, Jiang, Grosskopf and Berryman [9] discussed usingUnity as a virtual platform to explore HVAC systems in virtual worlds and present their research on educational constructs for achieving intended educational outcome goals using virtual worlds.

1.1.PKI Establishes SCS Student Chapter

In September 2010, PKI students interested in modeling and simulation (M&S) research proposed to establish a student chapter of the Society of Computer Simulation and Modeling (SCS) at PKI. In October 2010, a draft constitution and by-lawswere submitted for reviewto SCS Associate Vice President of Student Chapters, Professor Emeritus Tuncer Oren at theUniversity of Ottawa, Canada. The PKI SCS Student Chapter was approved by the SCS Board of Directors on January 31, 2011 thereby establishing the second SCS Student Chapter in the United States. Students elected to key positions included John Oerter, President; Matthew Morhardt, Vice President; Wyatt Suddarth, Secretary; and Jacob Partusch, Treasurer. In April 2011, Student Chapter President John Oerter attended the SCS Spring Simulation Conference in Boston, MA and was introduced to the symposium attendees during the Plenary Session. The SCS Student Chapter was an important initial step to attracting additional PKI students to M&S research which formed a critical mass of manpower and talent that attracted the interest of industry partners to work with PKI in this area.

1.2.Mutual Benefits of forming Research Partnerships

Industry companies and government agencies from various sectors such as architecture, architectural engineering,civil engineering, construction engineering and construction management, transportation and logistics and the US Strategic Command (USSTRATCOM) from Department of Defense all showed interest in collaborating with PKI on M&S research. Discussions with the companies and agencies were helpful to identifyinga number of mutual benefits to forming a research partnership.

Benefits to industry:

  • Enhance in-house research capabilities;
  • Teaming could potentially lead to new business and collaboration opportunities;
  • Creating and acquiring rights to intellectual property from research;
  • Avoid making expensive capital investments of company and agency funds into research infrastructure, technologies and laboratories;
  • Access to faculty expertise;
  • Student internships leading to possible new employee hires;
  • Possible leveraging of funds from other sources by jointly pursuing research grants.

Benefits to PKI faculty and students:

  • Access to real-world problems;
  • Exposure to project phases from concept development, engineering, construction and manufacturing;
  • Exposure to business and other ways of thinking;
  • Awareness and importance of costs, timelines and deliverables;
  • Creation of new course content for enriching academic programs;
  • Job opportunities for students through internships and hiring.

1.3.SAIC and PKI Research Partnership

Throughout this period, SAIC was the one entity more than any other that committed to forming a research partnership with PKI. This was primarily attributable to perseverance of two SAIC executives who championed this effort: Ms. Beverly Seay and Mr. Alfred Buckles. Ms. Seay was then Senior Vice President for SAIC’s Analysis, Simulation Systems Engineering and Training (ASSET) business unit and Mr. Buckles is the Vice President for Omaha Operations and the USSTRATCOM account manager. After discussing possible research areas, a formal teaming agreement between PKI and SAIC was put in place that gave PKI students access to SAIC modeling and simulation software to learn and gain experience in modeling and simulation. Additionally, SAIC and PKI agreed to jointly fund student support and development including year-round internships. SAIC also agreed to provide part time, on-site liaison at PKI which was filled by Mrs. Allison Baysa. Allison ensures students are provided with the proper equipment, software, instruction, training, information and other resources for them to conduct research and to ensure the program’s success.

2.Research objective

The primary goals of this research are to (1) develop a software system architecture that couples BIM, a virtual world simulation environment, and a gaming engine,and (2) develop a prototype software systemin which architects and engineers can design and develop 3-dimensional renderings of built environments. The development environmentsselected for this research wereAutodesk® Revit, Autodesk® 3DS Max [2], SAIC’s 3-dimensional virtual world calledOnline Interactive Virtual Environment (OLIVE) [8], versions 2.4.0 and 3.0, anda gaming engine called Unitydeveloped by Unity Technologies from San Francisco, California [10].

2.1.Specific Research Objectives

A major research objectiveof this project has been to create an intuitive, efficient and user-friendly virtual world tool that allows users to design and modify architectural objects rendered in BIM and then ported into a virtual simulation environment without losingengineering level specifications which will requireexact management of object attributes such as orientation, position, size and texture.

Another research objective is to give users the ability to customize attributes on-the-fly according to user or client preferences, constraints or prerequisites. Building designattributes such asinterior and exterior dimensions and finishes or the structural design of a building would be easily and quickly changedor modifiedwithout sacrificing high quality visualization, graphics and resolution of the rendered object.

A final key, high-level design objective is to provide developers with an asynchronous development and collaboration environment where simulationists, software engineers and system users will be able to collaborate online as individuals or groups from distributed locations.

2.2.Significance of Research to the Engineering and Academic Communities

Architectural and engineering firms frequently use BIM tools such as Autodesk® Revit, Autodesk® 3DS Maxand Bentley Building Applications, among others, to design and plan new buildings and for renovation projects ([2], [3]). BIM computer software programs enable architects and engineers to visualize new or renovated structures in 3Dspaces. These tools, however, are not well suited for making changes to existing architectural drawings or building designs that architectscould use to help their clients visualize newly designed or renovated spaces. Architectural teamstypically work around this problem by creating artistic renderings of the designed spaces for clients. The major drawbacks with artistic renderings are the expense and time required to create or modify them.

This research proposes a new architectural design environment that integratesBIM with a virtual simulation environment and gaming engine to give architects and engineers a virtual world in which to design spaces for clients with an intuitive,natural way in which the spaces can be explored and modified. A new virtual world BIM environment is intended to save architects, engineers, clients and other stakeholders’time and money by giving them the opportunity to make design changes and explore designed space functionality before construction.

In addition to serving architecture and architectural engineering industries, this research will also benefit academia and educational enterprises. Students studying architecture, architectural engineering, civil engineering, and construction will be well served by learning more effective ways to visualize and demonstrate important design concepts and processes.

Virtual environments can also provide engineers, construction workers, academic professors and students with opportunities to explore related topics such as construction site safetyand demonstrate how the construction team works together on projects without being exposed to unsafe conditions ([4], [5]). Real life site visits may not be possible due to access restrictions, cost, safety, and availability of students. Virtual environments will make it possible for teachers and students to examinevarious layers of building systems in ways that could not be done using building designs. Virtual classrooms will enable faculty and students to engage on personal and professional levels in ways that chat, texting, and cell phones will not support. Virtual worlds have the potential to profoundly make online learning more effective because students from all over the world can come together in acollaborative, virtual classroom to learn.

3.methoDology

SAIC’s simulation software package, OLIVE, served as the research and development environment for the project. OLIVE is currently used for applications in the areas of communications, networks, intelligent and embedded training, and modeling natural and man-made disasters. This research program was launched coincidentally at the same time that SAIC had undertaken a major effort to transition OLIVE from version 2.4.0 to 3.0. The transition involved re-architecting and re-hosting OLIVE to incorporate a new gaming engine, add new functionality and create a more robust toolkit for object development.

3.1.BIM Simulation System Prototype Architecture

The system architecture below gives a high-level depiction of the major components ofthe BIM simulation software system. The diagram also illustrates the generalprocess for how systemcomponents interact with one another to achieve design goals. Activities of system developers and users involved so far in the high-level design are also shown although the diagram does not necessarily portray the sequenceof steps in which they occur.

C Users user Desktop DiagramPicture PNG

Figure 3.1. High-level System Architecture

The diagram reflects the research team’s current understanding of client requirements for the prototype underdevelopment. It is expected that changes to the system’s architecture will occur as users add new requirements or modify the ways that components interactwith one another. The processes and interactions between clients, design team, data bases, OLIVE and BIM are based on use cases developed to date. As additional use cases are explored, lessons will be incorporated into the software system, component design and into the system’s architecture.

3.2.OLIVE 2.4.0

The team began its research and development by becoming familiar with OLIVE 2.4.0 functionally, as well as with methods and processes for importing customer designed objects into OLIVE using Autodesk® Revit and Autodesk® 3DS Max. Autodesk objects imported into OLIVE 2.4.0 must be modified using Autodesk® Revit and Autodesk® 3DS Maxtocomply with specific operating requirements of the OLIVE 2.4.0 simulation engine.

Importing an objectinto OLIVE 2.4.0,illustrated here using a building,begins with the instantiation of a structure in Autodesk® Revit which is then exported to Autodesk® 3DS Max as a .fbx file. In Autodesk® 3DS Max, each object must be manually taggedaccording to a hierarchical taxonomy for grouping and coupling objects. It is also necessary tomanually establishan object’sspatial references and geometric relationships between the imported object under consideration and other objects in OLIVE. Theseremaintedious, laborious and time consuming tasks.

Figure 3.2below illustrates a basic building rendered by the research team in Autodesk® Revit which took the team several hours to import into OLIVE 2.4.0.

Figure 3.2. OLIVE 2.4.0 Building

Research and analysisto improve OLIVE 2.4.0 functionality across several areas was conducted by the team during the past year. This effort led toa number of recommendations being acceptedaspossible improvementsto the processes outlined above and in OLIVE 3.0.

The teamanalyzed OLIVE’s collision mesh protocol that prevents avatars and other objects from passing through or colliding with each other or other objects. The collision mesh reference system uses up to 400 distinct vertices for rendering and positioning complex shapes [8]. Anotherresearch areawas related to the level of detail (LOD) mesh used for rendering object appearances. Typically, BIM object features are specified as attributes developed in Autodesk® 3DS Max. The upper bound for the LOD mesh is approximately 8000 vertices which gives developers a significant range for attribute specification [8]; something that also significantly increases object attribute complexity. Although theseremain manual processes, the team developed algorithms that simplified and streamlined the process of decomposing complex shapes into simpler objects with their own LODs and collision meshes.

Both the collision and LOD meshes must be parented to the master-rootwhich serves as an invisible (to the user) dummy object to which object attributes are anchored within the taxonomy. This establishes a coordinate reference position relative to the master-rootfor each componentthat forms theobject’s physical structure [8]. The specific attributes of each component, such as,materials, surfaces, textures and finishesmust be assigned in Autodesk® 3DS Max because the direct draw surface (.dds) extension is not supported by Autodesk® Revit. This requires objects to be imported from Autodesk® 3DS Max into OLIVE as .om or .model formats using appropriate plugins. The team’sanalysisof these areas led to development of new methods and guidelines for categorizing attributes and clarifying relationships between and among objectsusing Autodesk® 3DS Max. The research team’s new methods and processes streamlined the process of decomposing and incorporating complex objects into OLIVE.

3.3.OLIVE 3.0

OLIVE 3.0 does not yet offer a robust software development kit (SDK) of tools and plugins to allow users to add new or modify existing content to objects inthe virtual world. However, because it is built uponthe Unity gaming engine which supports many 3D file formats, such as,.3ds, .fbx, .dae and .obj, OLIVE 3.0could makethe process of importing objectsmuch smoother and less tedious.

As part of the joint SAIC-PKI research initiative, the research team undertook development ofnew OLIVE capabilities. Once the capabilities arefully developed and integrated into OLIVE 3.0,it will enable users to port objects from Autodesk® Revit into OLIVE without makingthe necessary intermediate adjustments or refinements in Autodesk® 3DS Max in order to port objects into the virtual world.

Figure 3.5 below illustratesabuilding ready to be exported intoOLIVE 3.0 that the team instantiated in Autodesk® 3DS Max. It was rendered, including edits, in less than 30 minutes. This illustrates a potential significant improvement ofOLIVE 3.0’s development environment over2.4.0 where it took the team several hoursto develop the same building.

Figure 3.5. OLIVE 3.0 Building

Finally, building OLIVE 3.0 on the Unity game engine platformmakes it widely and easily available as a web application with only a web browser required to view, edit and develop objects in OLIVE 3.0. This overcomes user limitations and restrictions which slow object development in OLIVE 2.4.0.

4.results and findings

Through the course of this project, the research team worked closely with architects and architectural engineers from academia and industry. The teamlearned a great dealabout the design needs and requirements of thesecommunities and, from this,identified potential solutions using the proposed BIM software system presented in this paper to meet many of those needs.

In working with the architectural and engineering communities and SAIC programmers, the team quickly found that the process of porting building information models into OLIVE 2.4.0 was far from ideal due to several factors. First, development time required for objects in Autodesk® 3DS Max is the most onerous and most difficult to overcome. Second, the burdensome requirement for OLIVE 2.4.0 files to be loaded ontouser computers as a precursor to collaboration significantly limited participation during both synchronous and asynchronous work sessions. OLIVE 3.0, on the other hand, offers a much improved platform with its improved graphics, web-based access, and easy-to-use object developmentenvironment.