Traffic conflict technique in the Czech Republic Ji Ambros - PDF document

24 th ICTCT workshop Traffic conflict technique in the Czech Republic Ji ř í Ambros Centrum dopravního výzkumu, v.v.i. (Transport Research Centre) Czech Republic Presentation 1) Introduction of TCT 2) History of TCT in the Czech Republic 3) Findings of literature survey 4) Pilot TCT observations 5) Lessons learned

Introduction of TCT (1/2) • tradition of using accident analyses for road safety assessment • drawbacks: regression to mean, quality of data, underreporting… ex post! • alternative approach: safety performance indicators including traffic conflicts (TCT) • TCT gives more data… ex ante! Introduction of TCT (2/2) • many TCTs around the world, general mutual agreement • TCT has to be reliable and valid • reliability ← training • validity ← correlation with accidents • they show the quality of TCT

Czech TCTs • usually known only at universities • Prague: manual observations • Ostrava: from video • both subjective Ostrava • not tested Prague • CDV in Brno Brno Literature survey (1/2) • focused on developing universal practical TCT method • TCT complements accident analyses • various opinions on validity / way to „market the procedure“ • progress to semiautomatic methods or video detection • do not develop but choose from existing ones

Literature survey (2/2) • survey findings show the way for Czech TCT: a) follow existing valid TCT method or b) use Czech method which has to be validated • need to study reliability and validity of both existing Czech TCT approaches • calibration by comparing their results • two pilots up to now… Pilot observations (1/3) 1) Brno: busy signalized intersection, mixed traffic (cars, public transport, pedestrians, cyclists) • 1 peak hour • two teams • manual vs video

Pilot observations (2/3) Conflict records were different – due to insufficient training before the observation? • mostly slight ones • severe conflicts were always recorded • safety diagnoses were generally the same Pilot observations (3/3) 2) Ostrava: non-signalized intersection, busy traffic, incl. tramway and students • 16 students, 3x 2-hour observations • repeated training by lecture • results were incomparable • some students improved • some lacked the motivation

Lessons learned • pilot observations show the need of training → reliability • lecture + testing observation + video e- learning for all target groups • then proceed to further observations • the aim is to produce common TCT guidelines for practical use 24 th ICTCT workshop Thank you for your attention Ji ř í Ambros Centrum dopravního výzkumu, v.v.i. (Transport Research Centre) Czech Republic