BSc thesis project proposal

Remote sensors and Internet of Things for rail-traffic applications

Project outside the university


At ProRail we are looking for opportunities of introducing an Internet Of Things approach for sensing along the rail-infrastructure. We are especially focusing on sensing the presence of a moving train from an industrial plant to a shunting yard. On shunting yards there is virtually no infrastructure for connectivity available (unlike a standard managed railway track with fiber optic, etc.) for signaling/safety systems with a high level of safety (Safety Integrity Level).
A typical business case at ProRail is being dictated by engineering costs (drawing revisions) when cables are involved, internal procedures and safety procedures for accessing to do any kind of installation on a yard. Our framework for innovation is that we want to stay away from rail-traffic safety-measures as maximum as possible. We would like to target for a totally different level of business case: the use of open source hardware like Arduino, Raspberry Pi, Intel IoT Galileo for the sensing seems an economic alternative. Also the expected launch of IoT networks like LoRa (KPN Netherlands, �) and SigFox will facilitate a lot of developments for the oncoming years.
It is not the cost of a sensor, but these other costs that dominate the business case and which we would like to eliminate. So any kind of external monitoring/sensing device which can just be bolted/clamped/fixed without any installation requirements, using wireless communication, which can run autonomously for a few years with batteries and additional energy harvesting with, e.g. solar or Peltier.
A particular challenge is the detection of trains using independent mini-sensor, e.g. based on Arduino or alike, with a some form of radar. Is it possible to come up with a sustainable design of a sensor configuration alongside a track with medium or even rare train possession? What detail of information can be produced when a train passes by at low speed: train presence yes/no, speed, or is it also possible to detect/count individual wagon wheels, detecting/counting wheel axis, counting wagons, even recognize form factors, visual signs and/or identification numbers? Can data be logged or is it only possible to act as a real-time monitor when a train passes and send out a signal before being forced back into sleep-mode in order to keep up a sufficient energy-level?
From our point of view, a balanced energy-management is the dominating factor to facilitate the design of a (combination of) sensor(s), the level of detailed measurement and processing capacity. At ProRail we would like to learn from design choices being made by students to come up with solutions for this challenge.


prof.dr. Alexander Yarovoy

Microwave Sensing, Signals and Systems Group

Department of Microelectronics

Last modified: 2017-03-28