Available online at www.sciencedirect.com
ScienceDirect
Procedia Engineering 00 (2016) 000–000
www.elsevier.com/locate/procedia
International Conference on Sustainable Design, Engineering and Construction
Optimizing BIM metadata manipulation using parametric tools A. M. Khaja, J. D. Seo, J. J. McArthur *
a
Department of Architectural Science, Ryerson University, 350 Victoria St, Toronto, ON, M5B 2K3, Canada
Abstract Building Information Management (BIM) is gaining popularity in the AEC industry for design and construction. Moving into the operations phase of the building, non-geometric information such as indoor environmental quality, and essential building services become increasingly important. The cost of maintaining the information in the BIM model has been a historical obstacle to adoption of BIM during this phase. To overcome this barrier, parametric tools such as Dynamo – often used for developing and manipulating model geometry – have been used to automate the information transfer between the BIM models and facilities management systems. This paper presents a case study investigating this technique, including series of investigations using parametric design tools, APIs and macros to classify, format, manipulate, and assign operations information to BIM elements. This case study demonstrates significant potential for automatic population of this non-geometric data into BIM models and the subsequent improvement for adoption of BIM in the operations phase. © 2016 The Authors. Published by Elsevier Ltd. Peer-review under responsibility of organizing committee of the International Conference on Sustainable Design, Engineering and Construction 2016. Keywords: Building Information Modeling (BIM), Facilities Management, parametric tools, building operations, visualization, analytics
1. Introduction The use of Building Information Management (BIM) models within the AEC industry for design, and construction is rapidly increasing, however there is a significant disconnect in BIM implementation within the operational phases of a building. Examining current BIM practice through a recent national survey of BIM practitioners saw only 12% of BIM users within the AEC industry pass on the model for use in management of
* Corresponding author. Tel.: +1-416-979-5000 x 4082. E-mail address:
[email protected] 1877-7058 © 2016 The Authors. Published by Elsevier Ltd. Peer-review under responsibility of organizing committee of the International Conference on Sustainable Design, Engineering and Construction 2016.
2
Khaja, Seo, McArthur / Procedia Engineering 00 (2016) 000–000
building [1] and less than 10% of BIM models used in building Operations [2]. BIM use for facilities management requires significant non-geometric data to support FM applications such as space management, comfort management, improved inventory management, and energy simulation and conservation [3],[4], as well as the information typically recorded in hard copies and manually entered into the Computer-Aided Facilities Management system(s) (CAFM). Although the benefits of BIM enabled facilities is widely documented [3-6], the cost and time resources and BIM proficiency required to both develop and maintain the BIM model [7],[8],[9] and challenges with data interoperability between the BIM model and existing CAFMs [7], [10] are significant barriers to adoption. To overcome these challenges, research is underway to explore how parametric tools (such as Autodesk’s Dynamo used in this research) can automate model maintenance and data transfer processes, thus reducing set-up and maintenance costs associated with the model. These tools are typically used for physical geometry manipulation within BIM models, however, researchers are beginning to look at the potential impacts of visual programming tools like Dynamo, for their ability to act as linkages for information exchange between the BIM model and external data sources [9],[11],[12]. The proposed approach uses the BIM as a “data consumer” as well as a Common Data Environment (CDE). Using a methodology based on common data formats maintains compatibility with CAFM and maximizes flexibility for future system expansion. The successful implementation of this automation reduces the time required for model upkeep and allows FM personnel to be proactive rather than reactive in maintenance and repair [5]. Given that operations and maintenance accounts for 60% of the overall project cost, the potential impact on time and cost through parametric tools is highly attractive to clients and owners [5]. This paper presents three investigations from a larger case study to develop a BIM in Sustainable Operations model of a campus building in order to demonstrate potential applications of the use of parametric tools such as Dynamo for managing data transfer between the CAFM and BIM CDE. This begins to develop a BIM-based standard approach to push CAFM data into the BIM database, recognized as a necessary development in the field of BIM in FM [13]. Note that while part of a larger case study, the focus of this paper is on the automation processes for data preparation and transfer; a concurrent publication discusses the FM information requirements used to develop – and use the results of – these investigations. Nomenclature BIM BIM-FM CAFM CDE FM HVAC HTML N PDF TSSA VBA .xls
Building Information Management BIM in Facilities Management model Computer-Aided Facilities Management Common Data Environment Facility Management Heating, ventilation and air-conditioning Hypertext markup language Number of owners sharing a space Portable document format Technical Standards and Safety Authority – the authority having jurisdiction on elevators in Canada Visual Basic for Applications Microsoft ExcelTM spreadsheet format (alternately, .xlsx)
2. Case Study Introduction A BIM for Facility Management (BIM-FM) project to develop a Virtual Campus Model has been under development for Ryerson University, starting with the Kerr Hall East (38% laboratories, 20% classroom, and 10% faculty offices) building, since 2014. This building consists of 201 rooms and other spaces (i.e. stairwells and corridors). This model was developed in Autodesk RevitTM 2015, and interfaces with Archibus – the primary CAFM software used for campus operations – along with several spreadsheet and web-hosted applications. The overarching objectives of this project are (1) to develop a virtual campus model consisting of BIM-FM models for all campus buildings nested within a larger site model, (2) to interface the component BIM models with CAFM data
Khaja, Seo, McArthur / Procedia Engineering 00 (2016) 000–000
3
and provide a single source of operations information for all campus buildings, (3) to explore potential BIM applications within the FM context and test their benefit in a real operations situation, and (4) to develop improved CAFM-BIM data transfer methodologies to overcome known barriers to adoption of BIM in FM. A guiding principle of the BIM-FM project is to work with existing FM processes and systems to minimize changes to operations processes (potential barriers to adoption), and automation tools are developed to support this principle. 2.1. Generic CAFM to BIM CDE Data Transfer Methodology Fig. 1 illustrates the data management process using parametric tools to mine data from the CAFM system and transfer it to the BIM CDE. To protect the integrity of the CAFM data, it is copied and exported into an intermediate format (in this case Microsoft Excel (.xls) format) prior to manipulation and mapping to the BIM model. Visual Basic for Applications (VBA) macros are used to sort, reformat, and consolidate and otherwise preprocess the data to align it with the shared parameters defined in the BIM model. Dynamo scripts map this preprocessed data onto the appropriate shared parameters and thus export it into the BIM CDE. The tabulated data is thus visualized in a 3D graphical format within the evolving BIM model of the facility.
Fig. 1. Generic CAFM to BIM CDE automation process indicating role of VBA macros and Dynamo scripts in data manipulation
This process was applied to three different investigations to explore and evaluate the potential impacts of parametric tool use to manipulate non-geometric building data, as presented in the following sections. 3. Investigation 1 – Space Management and Occupancy Tracking Automation This investigation explores the manipulation of qualitative metadata pertaining to physical spaces within Kerr Hall East. The term metadata here is used to refer to data describing the elements but not informing their geometry, and includes the ownership (e.g. Department code, department name, faculty name, and any sharing of spaces between them), functional group (division code), seating capacity, and signage information. Each of these, along with the room number code, which was used as the unique identifier for assigning these properties, was defined as a shared parameter within the model and attached to the room family. 3.1. Pre-Processing Because this data requires updates only after moves, renovations or as a result of new construction, no scheduling of updates was incorporated for the CAFM export, which was output in spreadsheet (.xlsx) format. Pre-processing of this data was critical because in the case of rooms shared between N owners, N replicate entries were exported with the same room number, and a VBA macro was required to delete the replicate entries and transfer the shared area assignments to new columns. 3.2. Data Transfer A Dynamo script (Fig. 2) imports the pre-processed CAFM output and associates qualitative metadata to the shared parameters matching the unique ID. Combined, the processing and data-transfer for 3000 data points took less than ten seconds and a quality assurance check verified zero errors arising from this transfer. These parameters are then displayed through view filters and summarized on the Rich Room Data Schedule (Fig. 3).
4
Khaja, Seo, McArthur / Procedia Engineering 00 (2016) 000–000
Fig. 2. Dynamo script used for data transfer into the BIM CDE
Fig. 3 Populated Rich Room Data Schedule indicating non-geometric data (partial view)
3.3. Results This approach was found to be very versatile, requiring little training or time to add new parameters to the model or data transfer process, and was quickly deployed with minimal resources, particularly compared to manual entry of the same data. By implementing view filters, visualization of the spaces by room type, ownership, or any other parameter, was quickly implemented, thus providing the facilities team with the means to visualize space and occupancy information in 3D. 3.4. Potential Process Improvement While a modification to the CAFM organization could optimize this process by removing the need for preprocessing need, this does not align with the guiding principles of the case study, and given the low frequency with which this script would be run, the cost of such a change would outweigh the benefit. SQL was also considered (with exports from the CAFM in an Access database format instead of Excel) but would complicate the remainder of the process as additional intermediate steps are currently required to import SQL data into Dynamo. 4. Investigation 2 – Work order tracking, visualization, and record-keeping This investigation explores the use of simple data analytics using parametric tools, the manipulation of large data imports, and visualization of consolidated within BIM in the form of a “heat map” to serve as a planning tool for
Khaja, Seo, McArthur / Procedia Engineering 00 (2016) 000–000
5
Facilities Managers. The CAFM work orders related to building maintenance tasks were used as a test case for this investigation, and the available data included 15M+ data points associated with 150,000 work orders created since 2010 cataloging all maintenance requests, occupant complaints, locksmith issues, custodial requests, and other operations and maintenance work requests, and both historical records and current (open) work orders were considered. These work orders originated from occupant requests via a web interface. 4.1. Pre-Processing In order to properly pre-process this data, historical data was exported from the CAFM into separate files for each year in order to both avoid data truncation and to reduce the file size for the most current work orders. A VFA macro was developed and used with each yearly export file to filter work orders by room, classify them by category, and develop summaries by month and room to map to the BIM CDE. Work orders were classified into categories (such as HVAC repair, general maintenance, interior finishes, custodial, etc.) by either searching the user input string for keywords (for open work orders) or the category assigned by the FM team during resolution (closed work orders). Fields associated with the shared parameters, including totals by year and month, Again, the room name was used as the unique ID for assigning the work order summary to the shared parameters for each space. 4.2. Data Transfer A Dynamo script (Fig. 4) relates the room name to the general maintenance schedule organized by the same unique identifier. To control the BIM model size while meeting the intent of this case study, only the monthly and annual totals in each category were included as shared parameters and view filters provided a simple means of visualizing this data within the BIM CDE. No scheduling is required for this historical data; the current work orders are updated daily by the FM team and thus this frequency is required for the automated export of current work orders from the CAFM system. 4.3. Results The use of Excel exports were problematic in that several files were required to export the data from the CAFM system in order to avoid data loss through file truncation – initially, whole campus data since 2010 (system deployment) was selected but only 2.5 years of work order data was written to the export file before reaching the Excel row limit. Changing to yearly reports solved this problem. To enable scaling to the multi-building context, the macro allows data to be filtered to include only the building(s) of interest for a particular iteration. The data processing and mapping to the BIM CDE were very quick (
