Machine Learning-Driven Predictive Analytics in Workload Distribution for Forensic Engineering Case Reporting Process: Dubai Police Case Study

The Dubai Police Forensic Engineering Department plays a significant role in forensic investigations, where efficient case reporting is essential for effective case resolution and the overall success of investigative outcomes. However, the department faces significant challenges in the workload distribution among examiners and reviewers, which directly leads to extended lead times and delays in case reporting. Therefore, this paper aims to address the challenge of unbalanced workload distribution of cases and its impact on case reporting by proposing a machine learning-driven solution, utilizing predictive analytics to facilitate case reporting, reduce lead times, and enhance the efficiency of forensic engineering operations based on the findings of an earlier study that identified significant inefficiencies within case reporting. By using historical case data, this paper applied predictive algorithms to identify patterns in case categories, incident types, and examiner allocation. Concept Explorer was used to generate variable combinations, while MATLAB was used to develop the predictive model that predicts the working days of different combinations. The findings demonstrate the potential of machine learning-driven solutions in achieving the department's goal of reducing average case reporting times to 7 days. The model reduced the working days for selected cases from an initial range of 1 ≤ WD ≤ 17 to a range of 6 ≤ WD ≤ 10, resulting in more balanced workloads and shorter lead times.