This study develops a proficient system econometrics framework targeted at optimising the reduction of fossil energy use and carbon emissions in the transportation industry, a case for KwaZulu-Natal Province of South Africa. By employing Mixed Integer Nonlinear Programming (MINLP) of the version Branch and Bound with Cuts, fulfilling Karush-Kuhn-Tucker (KKT) conditions, the research provides optimal solutions to balance energy consumption, emissions reduction and economic growth, following the impacts of other economic parameters on the panel datasets. The methodology integrates machine learning of Random Forest (RF) and Support Vector Machine (SVRM) with econometric regression techniques, enhancing the robustness and precision of the analysis using Gurobi Solver for the process implementation. Validation and sensitivity analyses were performed to confirm the robustness, accuracy and execution rate of the model. The results demonstrated high reliability with 96.5% precision stability. The results demonstrate that economic expansion and environmental sustainability are not mutually exclusive. Economic growth can be cushioned to grow by 43.2%, from 1,063.905 Billion ZAR in 2022 to 1,523.47 Billion ZAR to year 2035, while total energy consumption simultaneously declines by 12% (127.4 MJ to 112.3 MJ). In parallel, fossil fuel consumption decreases by 18% (3.50 Billion litres to 2.87 Billion litres), and \(CO_2\) -eq emissions decline by 22.5% (6.181 MT \(CO_2\) -eq to 4.79 MT \(CO_2\) -eq). This is to reflect the growing adoption of cleaner fuel alternatives, improved vehicle efficiency, and increasing electrification in the transportation industry. This demonstrates the potentials to carbon neutrality for sustainable transport solutions in the region. Globally, these findings support policies promoting clean energy, electric vehicles, and smart infrastructure to achieve carbon neutrality and sustainable transport.