Predictive user-based relocation through incentives in one-way car-sharing systems

• Design an innovative user-based relocation policy to solve the vehicle problem in oneway car-sharing systems. Integrate a Markovian prediction method forecast future demand losses due imbalances. Propose adaptive that determines optimal incentive based on customers’ value of time. Illustrate benefits and performance compared staff-based terms service level, cost sustainability. Describe hybrid operator-user-based Car-sharing systems are attractive alternative private vehicles their mobility However, distribution throughout network one-way is disturbed asymmetry stochasticity demand. As consequence, need be relocated maintain adequate level. In this paper, we develop approach through incentivization customers predictive model for state system Markov chains. Our methods determine as trade-off between expected omitted loss while taking into account time customers. We introduce learning algorithm allows operator estimate unknown customer preferences find incentive. Experimental results event-based simulation real show use incentives can significantly increase level profitability decrease number staff members needed balance system. Thereby, more sustainable relocations. Extensive sensitivity analyses prospective flexibility robustness our varying preferences.