1. - Using markers as spatial indices to create links
between physical locations and virtual information
stored in database
• Why not other indoor localization methods?
- Impossible for indoor environment (GPS)
- Time consuming calibration (WLAN, IMU) 1
- No accurate orientation information
Needed for AR/VR visualization
• Our marker-based MARvigator plugin solution:
- Marker: location ID saved in AprilTag enables devices to identify
specific regions in the environment
- Database: Geometry and facility information in BIM model
- Camera: Detect a marker to receive location information and
orientation of camera relative to the marker
Operation Manual for Users:
• Preparations phase (one-time process):
- Choose critical places to physically attach AprilTags
- Manually navigate to the virtual locations in BIM model and save the
location ID in a map file
• Operations phase:
- User needs to load the marker map file in BIM model and scan
markers
- The algorithm detects the marker in real-time
- The virtual camera of Navisworks receives the location ID saved in
markers and auto-adjusts to the corresponding locations in BIM model
• Achievement:
- Developed MARvigator plugin for Autodesk Navisworks, successfully
integrating indoor context-aware computing with BIM model
- Tested functions and potentials of Project Tango
• Limitation:
- MARvigator needs posting markers at critical locations. Owner would
prefer using a marker-less method for aesthetic purpose and easier
maintenance
• Next Steps:
- To develop a plugin on tablet to make MARvigator more portable,
since the plugin developed is for Navisworks 2015 on laptop
- Link BIM to Facility Management plugin developed by Walter P
Moore company, which saves facility maintenance information as a
visual bridge between CMMS and BIM models. This improvement
will enable the facility managers to extract useful information from
BIM models efficiently and easily adapt the plugin to specific needs
- Researchers can try to use doorplate or Google Project Tango as a
markerless method for easier maintenance, simultaneous positioning
and more efficient access to information
Professor SangHyun Lee, University of Michigan
Mr. Jim Jacobi, P.E., Walter P Moore company
[1] Khoury, H., Kamat, V. (2009). Indoor User Localization for Context-
Aware Information Retrieval in Construction Projects. Automation in
Construction, 18(4), 444-457
[2] Feng C, Kamat V R. Augmented reality markers as spatial indices for
indoor mobile AECFM applications. In Proceedings of 12th international
conference on construction applications of virtual reality (CONVR 2012).
2012: 235-24
[3] Project Tango. (2015). About Project Tango. Retrieved April 20, 2015,
from Project Tango: https://www.google.com/atap/project-tango/about-proj
ect-tango/index.html
CEE 530: Exploration of Context-Aware Mobile BIM
for Facility Management
Chen Feng1, Yingqi Liu2, Da Li2, Yuhang Xu2, Vineet R. Kamat1, Carol C. Menassa1
1Laboratory for Interactive Visualization in Engineering, University of Michigan, Ann Arbor, MI
2Construction Engineering and Management Program, University of Michigan, Ann Arbor, MI
Introduction
Methodology
MARvigator Plugin
Summary and Discussion
Acknowledgments
References
• Building Information Modeling (BIM)
- A 3D representation of a building’s information, including physical and
functional characteristics
- Mobile BIM aims to access BIM models on mobile devices
• Limitations of current mobile BIM applications
- BIM hasn’t been widely used in facility management
- Users find information via searching by names or manual navigation
Tedious and time-consuming process
Bad scalability for large sites or frequent periodic inspections
• Importance of context-awareness in mobile BIM
- Sensing the environment and reacting accordingly
- Indoor localization
Use users’ physical poses for auto-searching in BIM
Easy access to information of surrounding facilities
• Develop MARvigator plugin for Navisworks (Marker-based localization)
Figure 1: Interface
of MARvigator plugin
MARvigatorMonitorForm
MARvigatorPlugin
- Economical and efficient: localization in real-time; no energy
consuming infrastructure as in WLAN; easily scalable, etc.
Figure 2: Code architecture and general framework of MARvigator Plugin2
Figure 3: Location Recognition Using MARvigator
AprilTag
Further Research: Project Tango
• Project Tango:
- A context-aware tablet developed by
Google
- Featuring simultaneous localization and
mapping (SLAM)
• Exploring functions of Project Tango
- Depth Perception:
Rely on infrared light to detect the
outline of surrounding environment3
Measurement range: about 0.4-3.5
meters under optimal conditions
- Area Learning:
When user walks around, project
tango could record the trajectory and
capture all scenarios
Error self-correction: when the user is
exposed to the same location, project
tango would recognize the same
scene and localize the user
However: inaccurate absolute location,
accurate relative location
• Limitation of MARvigator: Discrete spatial Indices
Diagnostics
Camera
Image
Figure 4: Depth perception
of Project Tango
Figure 5: Area Learning
of Project Tango