StraWe – die STRAßenbahnWEiche / the tramway switch:
maintaining derailment safety and reducing noise in inner-city tramway traffic

03/2019 – 02/2022

Turnouts in light railway networks are typically inspected by trained personnel on a biannual basis. During these inspections, the switches are assigned a condition state from class A to D depending on criteria defined by the respective network operator, hence there is no national or international standardization of such assessments. In order to make these in-situ inspections more efficient, the project STRAWE aims to lay the foundation for an automated onboard monitoring technique for tram turnouts.

The goal is to assess rail wear at turnouts using a designated inspection tram which has been in operation in Vienna’s network since 2012. The light-sectioning equipment mounted onboard measures a rail profile for every centimeter travelled and computes the rail wear as the difference between the nominal rail profile and the currently measured one. The existing limits for wear of the rail head ensure good tracking of the vehicles in order to prevent derailment.

State of the art algorithms available for light-sectioning systems are only applicable for nominal rail profiles and cannot handle the variable cross-sections encountered around switches. Even for heavy rail systems, the automatic analysis of turnouts using light-sectioning is still in its early stages. For grooved rails on tram networks, no existing algorithms are known.

The ongoing work in the project ASSESS deals with the automatic detection of turnouts using noise and vibration sensors on an unpowered bogie of the inspection tram. Besides their detection, the project also aims to define a rating scheme for switches in terms of their relative loudness when driving across them. Given these existing results, the work proposed in STRAWE will continue from there by

  1. detecting switches and other rail head discontinuities using arbitrary instrumented light rail vehicles,
  2. automatically detecting switches using light-sectioning on the inspection tram, and
  3. combining the vibroacoustic data from 1) and the rail profile measurements from 2) to automatically obtain the integral condition state of a switch consistent with the in-situ inspections.

The work involves the temporary instrumentation of in-service vehicles in Vienna’s tram network such as the ultra-low floor (ULF) car and the new Flexity cars in order to produce a database of noise and vibration emissions of different vehicles as they drive across turnouts. The survey results will be compared to reference measurements of selected turnouts using the inspection tram (an old E1 car). This should provide information regarding the validity of onboard measurements when using arbitrary in-service vehicles and thus provide smaller network operators, who do not operate a designated inspection tram, with experience when it comes to equipping their in-service vehicles with sensors for monitoring rail condition.

By developing a new algorithm for light-sectioning on the inspection tram, optical and vibro-acoustic measurements can be combined to develop new indictors for assessing the rail wear at turnouts. Automatically assessing their condition when driving across them allows comparable, reproducible data to be collected for long-term studies of their degradation and ageing. Life-cycle analyses and in-situ inspections will benefit from this additional information by providing more specific, objective indicators than a discrete condition state between A and D. Furthermore, recommendations regarding the onboard assessment of turnouts can be provided to other Austrian tram network operators.

FFG Project 871516