The demand for exact mapping of railway equipment is growing so fast that conventional manual methods cannot keep up. A new industrial PhD will find an innovative solution to this.
Why let an employee travel hundreds of kilometres to register signs, signals, etc. along the lines when the work can be done faster, cheaper and more accurately with sensor technology?
In conjunction with the Department of Computer Science of the University of Copenhagen, Søren Andersen supervises industrial PhD student Georgios Karagiannis who will automate the process.
“We see a growing demand for mapping of the railway network, and we see a large potential for a digital solution, so that we can offer our clients a more efficient and accurate service. This will release working time for the employees that now gather data manually. Time they can use for other, more value-creating activities in COWI,” says Søren Andersen who like Georgios Karagiannis works in COWI Mapping.
A market worth billions of DKK
In the UK alone, operators have invested more than the equivalent of 100 MDKK in mapping of roads, utilities and railways, and according to COWI’s market surveys the expectation is that the European market will grow to the equivalent of several billion DKK in the coming years.
“The present markets consists primarily of large railway operators, and we expect the market to accelerate as operations are managed by more and larger organisations due to rationalisation and efficiency requirements. The necessary underlying systems will require more mapping with frequent updates, details and accuracy that manual processes cannot deliver,” says Market Director Michael Schultz Rasmussen.
From laboratory to real life
The automotive industry has already, for its driverless vehicles, developed technologies for real time collection and interpretation of data, but on the one hand they do not wish to share their knowledge, and on the other hand this technology has to be adapted to the railways.
Therefore, Georgios Karagiannis will first develop sensing and interpretation technology suitable for categorizing railway objects with precision. Then the data will then be compared with manually collected data, among them data available from the electrification of the Danish railways.
“When the technology is in place, it is crucial to take the models out of the lab and test them in real life to see how they react to fog, shadows and other changes in weather and lighting conditions. Later on we would like to invite select clients, internal and external, to join the project group so that they can participate in testing and fine-tuning the automated product before it is marketed,” says Søren Andersen.