This project was aimed at developing and validating a low-cost wireless worker proximity detection and alert sensing system for proactive safety warning in dynamic roadway work zones. A Bluetooth proximity sensing system was developed with customized software, which provided audio and vibratory alerts to both pedestrian workers and equipment operators simultaneously when they are in hazardous proximity situations through Bluetooth-enabled smart devices, such as smartphones and smartwatches. Extensive lab tests and field trials under controlled environments were conducted to develop and improve the functionalities of the proximity sensing and alert system. Furthermore, field tests were conducted at an earthmoving construction job site to evaluate the practicality of the system. From the test, the system performance was evaluated in real-world situations, and promising test results with positive feeback were obtained from the workers who participated in the test. The primary research findings include: (1) the Bluetooth proximity alert system provides reliable alerts during hazardous proximity situations, based on test results and feedback from workers who participated in the field tests; (2) the experimental results in controlled environments demonstrate that the Bluetooth proximity sensing and alert system provides reliable results with an appropriate alarm with slight performance differences when equipment approaches a worker at various speeds; (3) the adaptive signal processing (ASP) algorithm developed in this research was able to significantly reduce the signal processing delay and inconsistency of the Bluetooth system caused by vehicle’s approaching speeds; and (4) the field test results show that frequencies of hazardous proximity situations highly depend on the type of equipment and type of work to be performed nearby. The overall study demonstrates that the Bluetooth proximity alert system has high potential to promote safety in roadway construction due to its high accuracy, low cost, easy-to-use, scalability, and smart functions.