Andreas Lumbe Aas
NSB
Andreas Lumbe Aas holds a PhD and and an M.Sc. degree in information science from the Norwegian University of Science and Technology (NTNU). He is currently lead risk management, safety objectives and safety analysis in the dept. of safety at NSB, the Norwegian national railway operator. He has previously worked with RAMS for the Norwegian railway infrastrucure manager, as a scientist / reasearcher and as a risk and safety consultant within oil and gas.
Reserach areas currently in focus include SPADs, human performance and the implications of risk manangemet systems within large organizations. He works with cause analyses, risk analyses and security analyses. He is the system owner of the incident database Synergi Life at NSB, continously developing the system for use within the company, with focus on collecting incident data for analysis purposes.
Signals Passed At Danger (SPADs) is a safety concern for the railway industry due to the major accident potential. The majority of operators still use humans as safety barriers, handling restrictive signal aspects even though safety systems such as Automatic Train Protection (ATP) is widely used. Even with new technology, such as the ERTMS, human operators will dominate the next decades. This paper looks into some commonly accepted claims about SPADs.
Data in this study is from the Norwegian railway and includes 107 SPADs caused by human error in the years 2015 and 2016. Data from a company incident database was analysed. Safety professionals do structured interviews with train drivers after SPADs due to human error. This data set is the basis of this study.
The analysis revealed that 58 % of the drivers in the study had one or more SPADs prior to the SPAD in the study.
Claim 1: SPADs are more likely to occur on the last day before a holiday or day off. The analysis shows that 15 % of the SPADs occurred on the last day before a holiday or day off.
Claim 2: SPADs are more likely to occur on the first day before a holiday or day off. The analysis shows that 21 % of the SPADs occurred on the first day after a holiday or day off.
Claim 3: Less experienced drivers are more likely to have SPADs than more experienced train drivers. In this study, the average experience was 11.8 years and the mean experience was 6 years. Of the total, 57 % of the drivers had 6 years or less experience. 21 % of the drivers had two years or less experience. Looking at one year or less experience as a driver it was 17 %.
This study shows that one third of SPADs happen on the before or after a holiday or day off. Looking into the shift work of train drivers, this is a significant finding. Just above half of the drivers had shorter experience than the mean. This indicates that there is no significant overrepresentation between less experienced drivers, but a thorough analysis of all drivers’ experience remains.
Another interesting result was that the SPADs increased during the night, even though there are significantly fewer train movements at these hours. Data from 2012-2016 was then analysed (n=262) and confirmed the initial results and showed a peak from 22:00 to 01:00. One explanation for this increase could be that drivers are more tired during the night and thus less concentrated.
One threat to external validity is that the study comprise one single company in one country. A threat to internal validity is the precision of interviews.But the results are still valid considering these limitations.
The study supports claims on before/after holiday as well as driver experience. The findings in this study can contribute to targeted measures against the causes of SPADs, thus help reducing SPADs.
(Please consider this paper for a special edition of Applied Ergonomics).
Signals and signage; SPADs , Safety culture , Human error and human reliability
