Elsevier

Automation in Construction

Volume 48, December 2014, Pages 18-30
Automation in Construction

Natural markers for augmented reality-based indoor navigation and facility maintenance

https://doi.org/10.1016/j.autcon.2014.08.009Get rights and content

Highlights

  • Natural markers support AR-based indoor navigation and facility maintenance.

  • We evaluate the performance of natural markers for indoor navigation purposes.

  • We verify the practical potential based on case studies on the university campus.

Abstract

The longest phase in a facility's lifecycle is its maintenance period, during which operators perform activities to provide a comfortable living and working environment as well as to upkeep equipment to prevent functional failures. In current practice operators need a considerable amount of time to manually process dispersed and unformatted facility information to perform an actual task. Existing research approaches rely on expensive hardware infrastructure or use artificial, thus unesthetic Augmented Reality (AR) markers. In this paper we present a natural marker based AR framework that can digitally support facility maintenance (FM) operators when navigating to the FM item of interest and when actually performing the maintenance and repair actions. Marker detection performance experiments and case studies on our university campus indicate the feasibility and potential of natural markers for AR-based maintenance support.

Introduction

The longest period in the lifecycle of a building is the operation and maintenance (O&M) or facility management (FM) phase. In this phase, facility managers and operators perform activities to provide a comfortable living and working environment (e.g. pleasant temperatures) as well as to upkeep equipment to prevent functional failures. Since over 85% of the entire lifecycle costs are spent on facility management [1], improvements to the maintenance procedure will significantly reduce the overall building lifecycle budget.

Today's maintenance practice is characterized by dispersed and unformatted facility information that operators often need to manually browse, sort and select. Although software systems have recently been introduced, 50% of the on-site maintenance time is still spent on localizing inspection targets and navigating to them inside a facility [2]. Moreover, linked maintenance instructions are often multi-page documents, which sometimes are difficult to comprehend, in particular in case of emergencies.

Although some recent research studies propose to use Building Information Modeling (BIM) by either integrating or linking work order information to them, not all necessary information is currently available in a digitally integrated and standardized model. Moreover, available Ultra-wide Band (UWB), Wireless Local Area Networks (WLAN), Radio Frequency Identification (RFID) and Global Positioning System (GPS) indoor navigation approaches have been validated, but they rely on a costly equipment infrastructure for tags and readers. Existing Augmented Reality (AR) based solutions use artificial markers for both navigation and maintenance instruction support. This kind of marker is tedious to install all over a facility and also has some esthetical issues.

In this paper we propose a natural marker based Augmented Reality framework that can digitally support facility maintenance operators in performing their daily on-site maintenance jobs. Since 50% of the on-site maintenance time is still spent on localizing and navigating, and existing maintenance instructions are often multi-page, incomprehensible documents [2], our framework supports operators when (1) navigating to the FM item of interest and when (2) actually performing the maintenance and repair action. The main contribution of this paper is to highlight the big potential of natural markers, such as exit signs, to enable AR-based facility maintenance support. The presented methodology is implemented as a prototype and has been successfully tested on the university's campus. The results indicate the feasibility and the potential of the proposed framework.

Section snippets

Current practices

In today's maintenance and repair practice facility operators need to gather and access dispersed and unformatted facility information in order to handle work orders [3]. Typically, this information is handed over from the building design and the construction phase and is available in form of 2D drawings, spreadsheets, bar charts, field reports and paper-based guidelines. Collected in so-called Facility Document Repositories, the facility handover data is physically space consuming and might

Research framework

The overall project framework combines with Building Information Modeling (BIM) and natural markers to support AR-based facility maintenance activities (Fig. 2). By analogy with Ayer et al. [25], the proposed FM maintenance workflow is comprised of three major activities: (1) Digital Work Order (DWO) Compilation (collecting relevant information), (2) AR-based Indoor Navigation (positioning and navigation), and (3) AR-based Maintenance Instructions (performing maintenance task). While the first

Implementation

In order to test and illustrate the potential of the proposed framework, the methodology was prototypically implemented. For this purpose, an iPad 2 (processor: 1 GHz dual-core A5, max. camera resolution: 1280 × 720 pixels) and the Augmented Reality framework “metaioSDK 4.1.2” [26], including the picture marker tracking functionality, were utilized.

Fig. 5 depicts the software architecture and its major components: the metaioSDK framework, the mobile Operating System (OS) framework, positioning and

Conclusion and future work

The longest phase in a facility's lifecycle is its operation and maintenance period, during which facility operators perform activities to provide a comfortable living and working environment (e.g. pleasant temperature) as well as to upkeep equipment to prevent functional failures. In current practice operators need to manually process dispersed and unformatted facility information. Although software systems have recently been introduced, 50% of the on-site maintenance time is still spent on

Acknowledgments

The authors would like to thank Jens Neumann who contributed to this work in terms of preparing and running the experiments.

References (26)

Cited by (138)

  • Key functions in BIM-based AR platforms

    2023, Automation in Construction
  • BIM-based repair and maintenance for hospital work order management

    2022, Automation in Construction
    Citation Excerpt :

    Data collection for detecting and diagnosing target assets on-site has been a known issue in R&M activities [60]. Koch et al. [66] estimated a waste of 50% of the on-site maintenance time to identify targets and navigate inside a building. Several recent studies have employed AR technology to enhance navigation inside a building.

  • Augmented Reality for Maintenance Tasks with ChatGPT for Automated Text-To-Action

    2024, Journal of Construction Engineering and Management
View all citing articles on Scopus
1

Tel.: + 49 234 32 22109.

2

Tel.: + 49 234 32 23047.

3

Tel.: + 49 234 32 27009.

View full text