Quantifying the Benefits of Autonomous On-Demand Ride-Pooling: A Simulation Study for Munich, Germany

Abstract

On-Demand Ride-Pooling services have the potential to increase traffic efficiency compared to private vehicle trips by decreasing parking space needed and increasing vehicle occupancy due to higher vehicle utilization and shared trips, respectively. Thereby, an operator controls a fleet of vehicles that serve requested trips on-demand while trips can be shared. In this highly dynamic and stochastic setting, assymetric spatio-temporal request distributions can drive the system towards an imbalance between demand and supply when vehicles end up in regions with low demand. This imbalance would lead to low fleet utilization and high customer waiting times. This study proposes a novel rebalancing algorithm to predictively reposition idle fleet vehicles to reduce this imbalance. The algorithm first samples artificial requests from a predicted demand distribution and simulates future fleet states to identify supply shortages. An assignment problem is formulated that assigns repositioning trips considering multiple samples and forecast horizons. The algorithm is implemented in an agent-based simulation framework and compared to multiple state-of-the-art rebalancing algorithms. A case study for Chicago, Illinois shows the benefits of applying the repositioning strategy by increasing service rate and vehicle revenue hours by roughly 50% compared to a service without repositioning. It additionally outperforms all comparison algorithms by serving more customers, increasing the pooling efficiency and decreasing customer waiting time regardless of the forecasting method applied. As a trade-off, the computational time increases, but with a termination within a couple of seconds it is still applicable for large-scale real world instances.

Analytical and Agent-Based Model to Evaluate Ride-Pooling Impact Factors

“…Only a few parameters of different cities, such as surface area and average velocity, and the envisioned service quality parameters, such as detour time, maximum waiting time, and boarding time, are required. The model can estimate the shareability for different levels of market share without computationally expensive and data-exhaustive, but therefore more detailed and realistic, simulations ( 1 , 3 ).…”

“…In this section we describe on a high level the ODRP framework, which has the task to match vehicles and customers. For a detailed description of the algorithm and a comparison with other matching algorithms, we refer to Engelhardt et al ( 12 ).…”

“…On-demand ride-pooling (ODRP) services, in which trips with similar trajectories are shared, are considered to satisfy both the above-mentioned qualities. These services contribute to higher vehicle occupancy as a result of shared trips, and have the potential to reduce the number of vehicles and contribute to fewer vehicle kilometers driven in the city ( 1 – 3 ). However, the success factors of these services highly depend on customers’ willingness to share the ride with somebody else, which is highly influenced by personal preferences and service quality attributes ( 4 , 5 ).…”

Pre-Day Scheduling of Charging Processes in Mobility-on-Demand Systems Considering Electricity Price and Vehicle Utilization Forecasts

“…This case study at hand is based on data representing the city of Munich. The operating area (A x = 7 km and A y = 10 km), the trip rates λ i and trip lengths d i are based on a prior MoD study [22]. The average velocities per time interval are extracted from a traffic microsimulation model [23].…”

Network Fundamental Diagram Based Routing of Vehicle Fleets in Dynamic Traffic Simulations

“…This pre-processing step also costs some time, but only has to be done once as long as travel times remain constant. By pre-processing the network based on travel times from DTA for multiple time intervals, exogenous traffic congestion can be represented [13]. However, this approach cannot capture traffic dynamics generated by increased traffic efficiency due to pooling or induced demand from mode-choice decisions [14].…”