Predictive Positioning and Quality Of Service Ridesharing for Campus Mobility On Demand Systems

Justin Miller,Jonathan P. How
DOI: https://doi.org/10.48550/arXiv.1609.08116
2017-03-07
Abstract:Autonomous Mobility On Demand (MOD) systems can utilize fleet management strategies in order to provide a high customer quality of service (QoS). Previous works on autonomous MOD systems have developed methods for rebalancing single capacity vehicles, where QoS is maintained through large fleet sizing. This work focuses on MOD systems utilizing a small number of vehicles, such as those found on a campus, where additional vehicles cannot be introduced as demand for rides increases. A predictive positioning method is presented for improving customer QoS by identifying key locations to position the fleet in order to minimize expected customer wait time. Ridesharing is introduced as a means for improving customer QoS as arrival rates increase. However, with ridesharing perceived QoS is dependent on an often unknown customer preference. To address this challenge, a customer ratings model, which learns customer preference from a 5-star rating, is developed and incorporated directly into a ridesharing algorithm. The predictive positioning and ridesharing methods are applied to simulation of a real-world campus MOD system.A combined predictive positioning and ridesharing approach is shown to reduce customer service times by up to 29% and the customer ratings model is shown to provide the best overall MOD fleet management performance over a range of customer preferences.
Multiagent Systems
What problem does this paper attempt to address?
The problem that this paper attempts to solve is how to improve the Quality of Service (QoS) in the Mobility On Demand (MOD) system on campus through Predictive Positioning and Ridesharing strategies, especially when the number of vehicles is limited. Specifically, the paper focuses on the following points: 1. **Predictive Positioning**: In a MOD network where the number of vehicles is less than the number of customer demand points, a predictive positioning method is proposed to minimize the expected waiting time of customers by identifying key locations. 2. **Ridesharing**: As the demand rate increases, ridesharing is introduced as a method to improve the customer QoS. However, the perceived QoS of ridesharing depends on customer preferences which are usually unknown. 3. **Customer Rating Model**: A customer rating model for learning customer preferences from 5 - star ratings is developed and directly integrated into the ridesharing algorithm to address the above challenges. ### Main Contributions 1. **Predictive Positioning Method**: This method can minimize the expected waiting time of customers in a MOD network when the number of vehicles is less than the number of customer arrival locations. 2. **Customer Rating Model**: This model learns customer preferences from 5 - star ratings and serves as a customer QoS - oriented cost function. 3. **Scheduling - based Ridesharing Framework**: This framework can accommodate customer QoS indicators without encoding customer preference constraints. ### Problem Background - **Mobility On Demand (MOD)**: Provide on - demand travel services for customers through a shared fleet without the need for private vehicle ownership. - **Quality of Service (QoS)**: Factors affecting customer QoS include cost, comfort, safety, and convenience, etc. - **Self - management Strategy**: Existing self - managed MOD systems mainly optimize QoS through rebalancing strategies, but these strategies usually assume a large enough fleet, which is not always true in practical applications, especially in small - scale scenarios such as campuses. ### Solutions 1. **Predictive Positioning Algorithm**: - Use the known customer arrival rates to find the key nodes in the network graph so as to place unassigned vehicles at these nodes, thereby minimizing the expected waiting time of customers. - Select the best vehicle placement location by calculating the total waiting time and probability of each node. 2. **Ridesharing Algorithm**: - Use the insertion method as a heuristic solution to the ridesharing problem to assign new customer requests to vehicles to minimize the total QoS cost of the system. - Introduce a virtual "reject vehicle" to consider the situation of not serving customers, thereby improving the overall QoS. 3. **Customer Rating Model**: - Use the Random Forest classification decision - tree model to learn customer preferences from 5 - star ratings. - The trained Random Forest model serves as the ridesharing cost function, enabling the algorithm to adjust the service strategy according to customer feedback. ### Experimental Results - **Predictive Positioning**: At a low customer arrival rate, the predictive positioning method can reduce the customer service time by up to 20%. - **Ridesharing**: At a high customer arrival rate, the ridesharing method can utilize the extra capacity of vehicles to insert new customers into vehicle trips, thereby reducing the service time. - **Integrated Method**: The method that combines predictive positioning and ridesharing can significantly improve the customer QoS at all arrival rates, reducing the service time by up to 29%. ### Conclusion The paper demonstrates the effectiveness of predictive positioning and ridesharing strategies in improving the customer service quality in the MOD system. By combining these two methods, it is possible to effectively respond to different customer preferences and demand changes with limited vehicle resources, thereby providing a better travel experience.