Road construction causes a large number of fatalities and injuries among construction workers. A 2017 estimate from the National Institute for Occupational Safety and Health indicates that 60% of transportation-related fatalities occurred inside work zones from being stuck by a vehicle or mobile equipment. Therefore, there is a high need to prevent these struck-by accidents inside road construction zones. Many wireless tracking solutions [e.g., those based on bluetooth, radio frequency Identification (RFID) tags, ultra-wide band (UWB), and global positioning system (GPS)] have been proposed by the industry to better track the workers and equipment in the workzones to generate alerts when there is a risk of a worker getting struck by a vehicle or an equipment during construction. Although promising, these technologies have struggled to work robustly and accurately enough for real world large scale use. In this project, a robust UWB based wireless traction solution has been developed that allows UWB to work accurately even when there are obstructions, e.g., non-line-of-Sight(NLOS) situations) due to heavy equipment and vehicles in the area being tracked. The real-time UWB based location tracking and monitoring system called vehicle pose estimation using ultra-wide band radios (ViPER) developed in this project eliminates the effects of NLOS situations when they occur, thereby improving localization performance in those challenging scenarios. This project also designed and implemented a new localization approach and boundary estimation algorithm that accurately calculates the distance between tracked entities, e.g., construction vehicles and workers. Different configurations of ViPER were tested and evaluated in both for line-of-sight (LOS) and NLOS situations in lab settings using hardware prototypes. The system was finally evaluated in a road construction site to track and estimate the distance between vehicles, equipment and workers for accurate localization. The project developed hardware for location tracking, wireless sensor signal processing software program, and performed extensive in-lab and on-site evaluations.
The final report is available.