Assessment of the impact of driving behaviour of Motorcyclist on capacity of signalized intersections.

Abstract

This study assessed the impact of two key poor motorcyclist behaviours; queue‑jumping and red‑light running on the capacity and performance of two signalised intersections in Kampala namely; Nakulabye intersection and Upper Kololo Terrace/Wampeewo Avenue intersection. Traffic volumes and motorcycle maneuvers were extracted from video recordings during peak periods; for each signal cycle, counts of queue‑jumping and red‑light running events were recorded alongside start‑up and clearance lost times to enable computation of total lost time and effective green time , then saturation headway and saturation flow rate were calculated using Highway Capacity Manual procedures and intersections modelled with and without motorcycle interference in SIDRA Intersection software to evaluate their capacities, level‑of‑service and queue impacts . At the Nakulabye intersection, both queue‑jumping and red‑light running showed weak or negligible correlations with effective green time, total lost time, saturation headway, and saturation flow rate, suggesting that overlapping disruptions such as unauthorized minibus stops and constrained geometric conditions masked the effects of individual motorcycle violations. Conversely, at Upper Kololo Terrace/Wampeewo Avenue intersection, strong negative correlations were found between the poor behaviours and effective green time, and strong positive correlations with total lost time, despite limited visual evidence of severe traffic inefficiencies, indicating that the relatively stable traffic environment and lower background disruptions allowed even modest behavioural changes to produce consistently detectable impacts on capacity. These findings emphasize that the influence of poor motorcycle behaviour on intersection capacity is highly context‑specific; chaotic conditions may require holistic interventions addressing multiple sources of delay, whereas more orderly environments allow targeted behavioural controls to yield measurable improvements. It is recommended that future studies extend data collection across non‑peak hours, incorporate additional behaviours e.g., erratic acceleration, lane‑splitting, and tailor mitigation strategies to the unique operational characteristics of each site.