A new carpooling model and two-step solution approach for hybrid work environments

CO2Carpooling, a sharing economy model where groups share a car, is integral to daily life, offering economic efficiency and sustainability. In this paper, we study an extension of the long-term carpooling problem for hybrid work environments with individual working plans. We formulate the problem as a two-objective mixed-integer model considering total distance and fairness of used cars. Since the mathematical model cannot solve large-scale problems within a reasonable time, we develop a novel Two-Step solution algorithm (TSCBA). TSCBA is a population-based evolutionary metaheuristic combining distance-based clustering and local search. We test TSCBA using 24 problems generated through computational experiments. Moreover, a case study in Eskisehir, Türkiye, demonstrates the economic and environmental impacts of the proposed approach. Results show that weekly emissions and total weekly transportation cost are reduced by 12.64% and 37.91%, respectively, while car use is reduced by around 45%.