Seamus MacGearailt Profile Photo with Grey Background

Seamus MacGearailt

Technical Director
Transport Infrastructure Ireland (TII)


Transport Infrastructure Ireland (TII) has commissioned Roughan & O'Donovan (ROD) and ROD-IS to develop a methodology for the analysis of geometry-related collision risk on Ireland's national route road network.

The project aims to develop a practical framework for quantitatively identifying locations of greatest risk to enable better prioritisation of investment funding. 

Where realignment works are proposed, the model will provide a method for determining the likely impact of the proposed realignment on the safety of the route as a whole.


Research from a variety of countries, including the United States and Germany, has highlighted the importance of geometric consistency as a factor in collision risk. 

Realignment works to improve one section of roadway can, for example, have the undesired effect of increasing the collision rate at the interfaces between the improved section of road and the unimproved sections at either end.

A geometric design evaluation can be used to identify high-risk locations on the road network.  Improvement works and resources can then be concentrated on these sections to significantly improve safety performance.


ROD and ROD-IS have developed a dimensionless multicriteria model of collision risk that will consider a wide variety of road characteristics, including:

  • Horizontal alignment
  • Vertical alignment
  • Crossfall
  • Operating speed
  • Friction characteristics

The model will take into account the consistency of these factors along a given route.

The model is being calibrated with a data set derived from detailed surveys conducted on over 30 pilot sites.  

In parallel with this, using data previously collected during annual pavement surveys of the network, algorithms are being developed to derive horizontal and vertical geometric parameters for the entire single-carriageway portion of the national road network.

The dimensionless nature of the model will enable it to provide useful results, even in circumstances where the range of available data is limited.


Interested in our research projects? Learn more