Development of a Sensor Network Protocol for Non-invasive Detection of Mobile Plant Operation in Construction Sites Ruwini Edirisinghe 1 and Frank Boukamp 2 1) Vice Chancellor’s Research Fellow, School of Property, Construction and Project Management, RMIT University, Melbourne, Australia. Email: Ruwini.edirisighe@rmit.edu.au 2) Senior Lecturer, School of Property, Construction and Project Management, RMIT University, Melbourne, Australia. Email: frank.boukamp@rmit.edu.au Abstract: Tracking and monitoring techniques proposed for construction sites in the past have limitations and gaps. One is the privacy concerns associated with tagging and tracking of personnel or equipment on a site. Second is the practical limitation of tagging, as tagging the objects may not always be possible. Third is that tagging can lead to additional costs, as tags have to be maintained and repaired when damaged. To the best of authors’ knowledge, to date no non-invasive method has been developed that identifies on-site hazards in real time and links real-time hazard situations with building information models’ (BIM) site activity schedules. The authors aim to fill this gap through a relatively large project, of which the first phase work is presented in this paper. The proposed technique uses a wireless radio frequency based non-invasive detection technique to detect mobile plants in operation to identify risk exposure. It has the potential to determine risk profiles of ‘mobile plant in operation’ by detecting construction activity on site zones. This paper reports the development of the underlying IEEE 802.15 Zigbee based mesh network. The details related to the design and development of the communication protocol, sensor network configuration and preliminary testing are discussed. The network will be used on a construction site to detect the mobile plant operation non-invasively. This research is expected to contribute to improve safety of inherently dangerous construction industry, which globally reports relatively high accident rates compared to other industries. Keywords: wireless sensing, tag-free detection, tracking, safety, mobile equipment 1. INTRODUCTION Globally, construction is a dangerous industry. Unsurprisingly, due to the industry’s poor records in fatalities and serious compensation claims, the Australian Work Health and Safety Strategy 2012-2022 has identified the construction industry as one of the priority industries for safety improvement. Among the recorded fatalities, vehicle incidents accounted for 15% and being hit by moving objects accounted for 10% (Safe Work Australia 2012). A major reason for these incidents is the absence of a systematic mechanism to monitor risk exposure of ‘mobile plant’ on site. Definition of mobile plant from the work health and Safety Regulation 2011 (Work Health and Safety Act, 2011) is adopted in this study as ‘any plant that is provided with some form of self-propulsion that is ordinarily under the direct control of an operator, and includes: earthmoving machinery (e.g. rollers, graders, scrapers, bobcats), excavators, cranes, hoists, elevating work platforms, concrete placement booms, reach stackers and forklifts.’ This project proposes a novel approach of monitoring risk exposure on site. The proposed non-invasive mechanism monitors mobile plant in operation on site which will support safety improvements. 2. BACKGROUND 2.1 Hazard Recognition and Safety Planning Every resource on a construction site introduces its own set of safety hazards and is itself exposed to safety risks. Advances in the area of information and knowledge management allow for representation of and reasoning about construction information, for example by leveraging databases of project information, such as Building Information Models (BIM), to support site safety management. Sulankivi et.al. (2013) have reported on “integrating safety into BIM as an effective and practical method for detecting and eliminating fall-related hazards” prior to construction. Wang and Boukamp (2011) developed an ontology-based reasoning framework to support job safety analysis. Zhang et.al. (2012) later extended this framework and integrated it with Building Information Models to support “automated Jobsite Hazard Analysis (JHA) using safety ontologies in BIMs.” 1060