The role of predictive analytics in improving road safety in Zimbabwe

Abstract

This research was driven by the significance of road traffic accidents in Zimbabwe, leading to injuries, fatalities, and economic losses. The aim is to prevent unnecessary injuries and loss of life caused by road accidents in Zimbabwe through the need to leverage predictive analytics to identify high-risk areas and behaviors. The objective of the study is to assess the role of predictive analytics in improving road safety in Zimbabwe. The target population was all road users and stakeholders. The researcher used the number of registered vehicles in Zimbabwe (1,583,700, ZINARA 2023) to calculate a representative minimum sample size of 385 respondents using the Rao soft calculator (Cochran 1977 formula). The parameters used were a 95% confidence level (Z = 1.96), a 5% margin of error (d = 0.05), and a 50% response distribution (p = 0.5). In order to answer the research questions, the researcher employed a mixed-method approach through closed questionnaires (for quantitative data, easy to group and analyze) and key informant interviews (for qualitative, in-depth data and exploring context). The researcher focused on ensuring that the questionnaire is in line with the research’s objectives through matching every question to any one of the research’s objectives. The closed questionnaire used in the study was advantageous in gathering focused, quantitative data from respondents, which is easy to group, compare and analyze whilst key informant interviews in 15 conveniently sampled strategic organisations were used as a method of collecting more information that was not included in the closed questionnaires. Descriptive statistics were used to analyse the demographic profile of the survey’s respondents. A total of 391 participants responded to this survey with 118 (30.18%) being female, which is a fair representation of females considering the general gender bias in tech-related subjects. The participants also represented a broad age spectrum. In terms of educational attainment, the majority (77.50%) of respondents held a tertiary qualification whilst an overwhelming majority (96.42%) reported having a valid driver’s licence. The distribution also suggests that most respondents were seasoned drivers, likely with substantial exposure to road traffic conditions and safety issues. Further analysis of the results indicated that predictive analytics was indeed effective in improving road safety with a significant majority of respondents (65.98%) believing that predictive analytics can reduce road accidents, and 97% agreed it can identify high-risk areas. However, challenges were noted, including resource constraints, a lack of stakeholder collaboration, and limited staff technical abilities. The major causes of accidents categorized into five domains which where human-related factors, vehicle condition, environmental factors, governance/enforcement as well as socio-economic factors. Historical accident data was identified as the most crucial data source, followed by driver behavior, traffic flow data, and vehicle data. There is also evidence that predictive analytics will help with resource allocation and enhance decision-making and policy formulation for transport authorities. Several positive outcomes have already been influenced by data, such as targeted awareness campaigns, warning sign installations, and major infrastructure projects. While the general understanding of predictive analytics is high (88.49%), awareness of its application specifically in road safety is lower. There was a statistically significant association between educational attainment and understanding of predictive analytics (χ²(8) = 53.01, p < 0.001). The research conclusion is that predictive analytics has significant potential to transform road safety in Zimbabwe by shifting from a traditional reactive approach to data-driven prevention strategies. For successful adoption, the study recommends a multi-pronged approach which includes strengthening data infrastructure, building institutional and technical capacity, support multi-sector collaboration, deploy technology-driven Interventions, implement public awareness strategies as well as allocate sustainable funding.