Ad-hoc train-arrival notification system for railway safety in remote areas

Aida Eduard, Dnislam Urazayev, Aruzhan Sabyrbek, Daniil Orel, Dimitrios Zorbas

Research output: Contribution to journalArticlepeer-review


In the last few years, a particular interest in using wireless technologies in the industrial domain in order to automate processes and increase the level of safety has been noticed. This paper introduces an affordable mobile system to notify railway workers in rural areas about the approach of trains, and thus, to enhance their safety allowing for the early evacuation of repair sites located near the rails. The system comprises three key elements: a train device, a portable station device, and wearable devices for the workers. The communication methods and the underlying protocols between these components are discussed in detail. The system has been developed for freight trains of the national railway company of Kazakhstan and has undergone extensive testing for each of its components before its final trial. The preliminary results demonstrate that the system meets the requirements in terms of evacuation time, range, and portability, while exhibiting a very low cost of manufacturing. More specifically, the system can achieve a reliable communication range of several kilometers and a maximum response time of 2.3 s. The cost does not exceed $500 for a set of train, station, and 5 worker devices.

Original languageEnglish
Article number101062
JournalInternet of Things (Netherlands)
Publication statusPublished - Apr 2024


  • IEEE802.11
  • Internet of Things
  • LoRa
  • Machine Learning
  • Railway safety

ASJC Scopus subject areas

  • Software
  • Computer Science (miscellaneous)
  • Information Systems
  • Engineering (miscellaneous)
  • Hardware and Architecture
  • Computer Science Applications
  • Artificial Intelligence
  • Management of Technology and Innovation


Dive into the research topics of 'Ad-hoc train-arrival notification system for railway safety in remote areas'. Together they form a unique fingerprint.

Cite this