Growing traffic on US roadways and heavy construction machinery on road construction sites pose a critical safety threat to construction workers. Each year over 20,000 workers are injured on US work-zones and more than 100 workers lose their lives in work-zone accidents. Most of these injuries occur due to accidents by construction vehicles (35%) or due to accidents involving passing-by traffic (12%), and almost all work-zone fatalities occur due to these two kinds of accidents (1). One way to reduce the number of such accidents is to alert the operators of the heavy construction vehicles as well as passing-by drivers about workers’ presence in and around the work-zone. This project’s main target is to improve workers’ safety with a secondary goal of enhancing traffic mobility on roads with work-zones. We have addressed the workers’ safety issue especially for workers who work around heavy machinery by proposing and developing a dedicated short range communication (DSRC)-based safety/warning system, which can provide visual guidance to the operators of the construction vehicles about the workers’ locations in the vicinity. This visual guidance keeps the operators of the heavy machinery well informed of the workers’ whereabouts in close proximity. In addition to enhanced worker safety, our proposed system can dynamically post suitable speed limits and/or warning messages on DSRC-equipped variable message signs (VMSs) to warn the drivers of the incoming vehicles about the presence of the workers in an active work-zone. This helps improve traffic mobility and driver vigilance while eliminating the need for manual input to VMSs. This project aims to improve worker safety by providing visual guidance to construction vehicle operators. To achieve this goal, the location information of workers with respect to the construction vehicle needs to be acquired and communicated to the operator in a useful manner. For this project, we have used the Global Positioning System (GPS) to obtain the location information because this method of acquiring position is much less dependent on line of sight (LoS) in comparison to other methods such as sensors. In addition to this, GPS needs much less processing power compared to other methods such as video-based positioning systems. To transmit this information from the workers to the construction vehicle, any kind of wireless communication technology can be used. In this project, we have used DSRC based communication primarily because with sufficient DSRC market penetration rate in the future, the proposed warning/safety system can be expanded to the DSRC-equipped vehicles passing by. Furthermore, DSRC offers high security and unaffected performance in severe weather conditions (2). The proposed system requires each worker to carry a device containing a GPS receiver and a DSRC radio to periodically transmit their location information to nearby construction vehicles. A study was conducted in collaboration with a DSRC device manufacturer, Savari Inc., to determine the feasibility of developing such a wearable device for workers. The results of this study show that a small device with a GPS receiver and a DSRC radio can be developed to be embedded in a worker’s safety vest. The study also showed that it would be economically feasible to have this wearable device manufactured in large quantities. The worker’s device will be referred to as wearable safety device (WSD), for the rest of this report. The system also requires the construction vehicle to be equipped with a DSRC radio to receive the location information from the nearby workers and with a monitor screen to graphically display this location information to the construction vehicle operator. The monitor screen needs to communicate with the on-board DSRC device to obtain the information needed for graphical display. The secondary goal of the project is to enhance traffic mobility in work-zones by alerting the drivers of passing vehicles about the presence of workers via VMSs. To achieve this goal, a VMS needs to be equipped with a DSRC device to receive the locations of the workers, if present, so that it can prepare the VMS to display an appropriate message, e.g., a warning sign and/or a speed limit depending on the presence of workers. A crucial issue with using GPS receivers for safety applications is that the position accuracy of consumer-grade available GPS receivers is not high enough. The regular GPS device has a position accuracy in the range of 3-5m. There are some specially designed GPS receivers, e.g., differential GPS receivers, that can provide very high position accuracy, up to a few centimeters. However, such devices are costly. In this project, each worker needs to have a WSD with a GPS receiver so using expensive GPS receivers will make the WSD cost-prohibitive. Therefore, in this project we have used regular GPS receivers to achieve acceptable relative position accuracy based on differential GPS principle. The differential GPS principle is used in differential GPS (DGPS) receivers, which are very accurate but expensive. In the DGPS method, there are fixed ground-based reference stations for which the GPS coordinates are known. The GPS data acquired from the satellite signals is compared to this set of known location data and the amount of error in the GPS reading is determined. This correction is then transmitted via Radio Frequency (RF) communication to the nearby DGPS receivers to correct their GPS readings before final positions of the receivers are calculated. This improves the accuracy because the biggest part of GPS error is due to atmospheric disturbances and remains the same over a large geographical area, and DGPS receivers can use the same correction as the ground-based stations. In this project, we have chosen to use a similar approach. Since a bigger part of GPS error remains the same over a large area, the GPS receivers of the worker and the construction vehicle will have a similar atmospheric error. We show the construction vehicle operator the relative GPS locations of the workers with respect to the location of the construction vehicle (as opposed to absolute locations of the workers) which essentially eliminates the larger portion of GPS error. The proposed system has three main components: The WSD, the DSRC-equipped construction vehicle with on-board monitor screen, and the DSRC-equipped VMS. The development progress of each component is described below: Wearable Safety Device: We have not yet developed the WSD at this stage but we have developed the software application for the WSD to communicate with the on-board DSRC device of a construction vehicle. We have tested this software application using a regular dashboard DSRC device, which also has a built-in GPS receiver. The DSRC/monitor-equipped Construction Vehicle: We have developed the software for the DSRC device on-board of the construction vehicle to acquire the location information of the workers. We have also developed the interface for the on-board DSRC device to communicate workers’ acquired positions to a monitor screen. We have used an Android-based tablet as the monitor screen that communicates with the on-board DSRC device via a local wireless connection using User Datagram Protocol (UDP)- based communication. The information about the location of the workers and the construction vehicle relative to each other is sent from the on-board DSRC device to the tablet and the tablet displays this information to the construction vehicle operator using the application software, which we have developed for this purpose. The DSRC-equipped Variable Message Sign: We have developed the software for the DSRC-equipped VMS to auto-detect the presence of the workers and change the posted message on the VMS, e.g., speed limit or another warning sign. The hardware interface between the DSRC device and the VMS is serial communication. The proposed system has been developed for the proof of concept evaluation, and we have conducted field tests to demonstrate functionality and feasibility of the system. The test results show that the system displays workers’ positions on an Android-based tablet with an acceptable distance accuracy of 1.5 -2 m. We have achieved a direction accuracy of 15 — 20 degrees, we are also working on a different method to improve the direction accuracy, this method has resulted in a much-improved accuracy (up to 5 degrees) in our preliminary tests. This level of distance and direction accuracy is sufficient for the operator of a construction vehicle to correctly identify the locations of the workers. We have also confirmed that the system is capable of detecting multiple workers at the same time. Furthermore, the functionality of the variable message sign was tested in a lab environment, and our test results show that the DSRC-equipped VMS can auto-detect the presence of a worker in its vicinity and post a proper speed limit and warning message without any kind of manual input. (Author/publisher)
Samenvatting