Railroad environments are generally considered to be among the most dynamic workplace environments, even with constant improvement efforts by the railroad industry. While there has been great progress in equipment safety, personnel safety is a significantly harder challenge. These challenges are primarily derived from the presence of heavy moving machinery in close proximity to personnel and the difficulty of designing reliable wearable protection devices. Additionally, variable weather conditions, challenging walking conditions (ballast, trip hazards, etc.), and difficulty to focus on environment, moving objects, and on tasks at hand place the employees in constant peril. Therefore, our survey is focused on exploring solutions for protecting employees through unified system modeling and design that makes the employee integral to the process and results in personal protective devices that work with the environment and the employee, not against them. The optimal system design integrates not only protection of the employees from falls, unsafe practices, or collisions, but also aids in resource planning, safe operation and accounting of “near-miss” situations.
In recent years the railroads have made significant investments in process automation and monitoring solutions such as Wireless Sensor Networks. These technologies are becoming increasingly cloud-connected and autonomous. They provide a plethora of information about equipment positions, movement, railcar lading, and many other factors, all of which are highly useful in the design and implementation of a railyard worker protection system. They allow us to predict position and movement, and can thus be used to provide effective proximity detection and alerting in some railyard regions where these systems are installed. Additionally, we discuss several technologies addressing near-collision, fall, and proximity situations through RF and non-RF-based techniques. The railroad industry has been advancing efforts leveraging these technologies to improve the safety of their workers.
However, there are also many challenges that remain largely unaddressed. For example, in railroads, a detailed and exhaustive causation analysis for worker incidents has yet to be conducted. Therefore, in an environment like a railyard there is no solution to detect or prevent Employee on Duty (EOD) fall, collision, or health issues such as dehydration, psychological issues and high blood pressure. Protective devices worn by workers is believed to be one of the most important, cost-effective, and scalable potential candidate solutions. Recent advances are making wearable wireless body area networks (WBAN) and wireless sensor networks (WSNs) that are distributed and large-scale a reality. Such distributed networks consist of wearable sensors, fixed-installation sensors and communication links between all of them. The challenges are found in selecting wearable sensors, researching reliable communication among nodes without interfering with proximity detection and suitable for high-multipath, non-line of sight channel conditions, wearable antenna designs, power supply requirements, etc.
A dense, distributed, large-scale environment like a railyard requires comprehensive workspace modelling and safety analysis. Interference related to RF sensor deployment, blind spots in vision-based approaches, and wireless propagation in intra and inter-WBAN communication due to dense non-Line-of-Sight workspace environments, metallic heavy machinery and the use of RF sensors, are all individual research challenges in this domain.
This paper reviews these challenges, explores potential solutions, and thus provides a comprehensive survey of a holistic system design approach for a wearable railyard worker protection system that is unobtrusive, effective, and reliable.
A mobile localization method based on a robust extend Kalman filter and improved M-estimation in Internet of things
