Slide 1

Slide 2

Investigating Pedestrian Injury Crashes on Modern Roundabouts in Addis Ababa, Ethiopia Getu Segni Tulu, Md. Mazharul Haque, Simon Washington, Mark King 1

Slide 3

Outline Background Methods Result Discussion and Conclusions 2

Slide 4

Background 3 -Roundabout introduced 40 years before (1) -Modern roundabouts have four main features

Slide 5

Background (cntd) Roundabouts reduce the number of crashes and injury severity level Safety performance of roundabouts in developing countries not known Many differences in roadway, traffic, and road users behavior 65% of vehicles are more than 15 years old 40% of total fatal crashes in Africa pedestrians (2) 85% of total injury crashes are pedestrians (3,4) 4

Slide 6

Background (cntd) Objective of the Research to develop crash prediction models of pedestrian crashes at roundabouts using geometric, traffic and land use characteristics affecting pedestrian safety to propose potential remedial measures 5

Slide 7

Methods Data collection Four categories of data collected from 22 modern roundabouts three years pedestrian crashes exposure variables: ADT, Pedestrian crossing volume roadway and traffic variables number of lanes, roundabout configuration (four-legged or three-legged), posted speed limit, maximum gradient of the approach, presence of a concrete barrier as a divider, 6

Slide 8

Methods (cntd) presence of a bus stop within 50m of the yield line on an approach road presence of raised medians, presence of a footpath, presence of a fence along the footpath spatial Variables/ site characteristics school zone within 100m of the roundabout, availability of alcohol-serving bars within a 300m radius of the roundabout, land use (roundabout was located in office and commercial areas), and the presence of a public transport terminal beside the roundabout 7

Slide 9

Methods (cntd) develop the pedestrian safety performance function using count model Random effect models (6,7) Random effect Poisson Model (REP) Random effect Negative Binomial model(RENP) Model Evaluation and Selection Criteria (8,9) 8

Slide 10

Result: Random Effect Poisson Model 9 VariablesCoff.IRRStd. Err ZP>|z|[95% Conf. Interval] Constant -9.71*** 3.23-3.010.00[-16.04, -3.39] Log of the product of major road and minor road AADT (aadt1 aadt2) 0.82**2.260.392.100.04[0.06, 1.58] Log of average daily pedestrian crossing volume 0.65**1.920.322.050.04[0.03, 1.28] Maximum gradient of major approach road 0.11***1.120.042.950.00[0.04, 0.19] Presence of raised medians on approach road -0.41**0.660.18-2.290.02[-0.77, -0.06] Availability of terminals beside the roundabout 0.50***1.650.202.530.01[0.11, 0.89] Presence of schools at 100 meter radius -0.38*0.680.20-1.870.06[-0.78,-0.02] Note:***,**,* significance at 1%, 5%, 10% level respectively

Slide 11

Discussion & Conclusion REP model is marginally superior RENB model in explaining pedestrian crashes two exposure variables the product of major road and minor road AADT and pedestrian crossing volume design proper pedestrian crossing infrastructure 10

Slide 12

Discussion & Conclusion (cntd) Provision of raised median along major roads Gradient of approach road and stopping sight distance should be properly designed Special intensity pedestrian exposure when roundabout near to public transport terminals Entry and exist of pedestrians and vehicles far from roundabout influence areas reduce the vehicle-vehicle and pedestrians conflicts Equips sidewalks with pedestrian fence to reduce illegal crossing 11

Slide 13

Discussion & Conclusion (cntd) 12 Typical modern roundabouts in Addis Ababa Suggested improvement on existing modern roundabouts

Slide 14

References 1.Solomon, T.G., Capacity evaluation of roundabout junctions in Addis Ababa, in Civil Engineering Department. 2007, Addis Ababa University Addis Ababa: Addis Ababa 2.WHO. (2013b). Pedestrian safety: a road safety manual for decision-makers and practitioner 3. Downing, A., G. Jacobs, A. Aeron-Thomas, and J. Sharples, Review of Road Safety in Urban Areas. 2000, Transport Research Laboratory and Department for International Development: London. 4.Jacobs, G., A. Aeron-Thomas, and A. Astrop, Estimating global road fatalities. 2000, Transport Research Laboratory and Department for International Development: London. 6. Washington, S.P., M.G. Karlaftis, and F.L. Mannering, Statistical and Econometric Methods for Transportation Data Analysis. 2011, New York: Taylor & Francis Group. 7.Lord, D. and F. Mannering, The statistical analysis of crash-frequency data: A review and assessment of methodological alternatives. Transportation Research Part A: Policy and Practice, 2010. 44(5): p. 291-305. 8. Akaike, H. Information Theory and an Extension of the Maximum Likelihood Principle. in Second International Symposium on Information Theory. 1972. Budapest Akadmiai Kiad. 9. Schwarz, G., Estimating the dimension of a model. The Annals of Statistics, 1978. 6: p. 461-464. 13

Slide 15

14 Thank you Ethiopian Roads Authority (Ato Zaid W/gebrial, Ato Bekele Nigussie, Ato Alemayehu Ayele) Ato Tatek Abera Rama Consulting Engineers Ato Efrem G/ Egziabher Lidet consultants Ato Andualem Haile Georgis STADIA Consultant