Date: April 14, 2008
Time: 10:30am
Location: NIA, Room 137
Additional Information: Presentation (PDF)

The purpose of this talk is to apply the Causal Model of Air Transport Safety (CATS) to assess the risk benefits and risk penalties implied by NASA’s new Merging and Spacing concept. The goal of the new concept is to increase runway throughput. Increasing throughput without a corresponding reduction in the accident probability per flight would result in a higher accident incidence, and higher perceptions of air transport risk. A system wide risk model, like the CATS model, can be employed to quantify the risk consequences of the new merging and spacing concept.

CATS was commissioned by the Dutch Ministry of Transport and is being developed by a consortium including Delft University of Technology, National Aerospace Laboratory (NLR), Det Norske Veritas (DNV) and White Queen. The project arose from a need for a thorough understanding of the causal factors underlying the risks of air transport, so that efforts to improve safety can be made as effective as possible. This can be achieved by developing a fully operational causal model that represents the causes of commercial air transport accidents and the safeguards that are in place to prevent them. The underlying software tool, UNINET, was developed at the Delft Institute of Applied Mathematics of the Delft University of Technology to enable the mathematical representation of the model through a powerful mathematical tool, the Bayesian Belief Nets (BBNs). UNINET is used for dealing with large continuous-discrete non-parametric BBNs, but model learning may also be performed in this standalone software package.