Assessing the Effect of Demand Shift on Fully Shared Autonomous Transportation System.pptx

Demand Shift on Fully Shared Autonomous Transportation System

1. Assessing the Effect of Demand Shift on Fully Shared Autonomous Transportation System Assessing the Effect of Demand Shift on Fully Shared Autonomous Transportation System

2. From  early studies in literature, for fully Shared Autonomous Vehicle (SAV) System required Fleet size approx. 1/10, Demand is confined in region ​ What if substantial shift in demand from the region to outside (i.e. vacation or any other reasons)? ​ 2 Background of Study Ciari , Francesco, Maxim Janzen, and Cezary   Ziemlicki . "Planning shared automated vehicle fleets: Specific modeling requirements and concepts to address them."  Demand for Emerging Transportation Systems . Elsevier, 2020. 151-168.​ Assessing the Effect of Demand Shift on Fully Shared Autonomous Transportation System Original Network Vacation location

3. Many studies are conducted to calculate fleet size considering average day traffic data, but none of them is focusing on scenarios with deterministic and non-deterministic fluctuations in demand.​ Level of service is compromised and in worst case could not fulfill the demand ( measured by waiting time, turnaround time)​ System studied in isolation, excluded origin and destination of outside trips.​ Propagation effect : need of multi day simulation with different scenarios​ Intended Research Direction: ​ What should be fleet size or operation strategies given a level of service with demand fluctuations? 3 Research Gap Assessing the Effect of Demand Shift on Fully Shared Autonomous Transportation System

4. Base Scenario: Input Data: Sioux Fall Network , Initial Plan/ Population Endogenous Variable: waiting time, Service time (waiting time + travel time) Exogenous Variable: Different level of Vehicle Supply Demand Shifted Scenario (Holiday/Vacation): Input Data: Sioux Fall Network Endogenous Variable: waiting time, Exogenous Variable: Different level of Vehicle Supply (Under Modified Plan by removing agents), Modified Plan/ Population (Demand Shift) What strategy to be adopted for Modified Population & Vehicle Detouring? 4 Methodological Framework Assessing the Effect of Demand Shift on Fully Shared Autonomous Transportation System

5. Strategy for modification of Population file: Agent Removed: Agents will be removed randomly maintaining some spatial-temporal distribution [Different level of demand shift will be considered 5%, 10%, 15%, …] Vehicle Strategy_1: Agents takes the vehicle with them and vehicles doesn’t return, therefore vehicle also removed Vehicle Strategy_2: Vehicles returns after dropping off the passenger Agents Returns back Vehicle Strategy_1: Agents takes the vehicle with them and vehicles stays with them and returns with them vehicle reappears. Vehicle Strategy_2: Vehicles goes again to pickup passengers Modified Population & Vehicle Detour Strategy 5 Assessing the Effect of Demand Shift on Fully Shared Autonomous Transportation System

6. Base Scenario:   The initial scenario is generated with average day plan for each consecutive day for a whole week.  Input: Sioux Falls plan file: Either same plan for each day, or modified plan (to be decided), vehicle density  AV Vehicles: Provide randomly and check the performance on waiting time.  Results and graphs:  With same level of supply/vehicle, what happens to the waiting/service time in base network location.  Level of service: and waiting time, VKT  Holiday scenario:   Remove agents from plan as well as vehicle and then compare level of service with base scenario  Agent removal: Random, not focusing on holiday location (spatial-temporal analysis). The analysis is limited to study area.  Logic for vehicle back to city: Random or distance based, or facility at destination location ( not analysed right now, study is limited to city level (Origin location))  Different levels of demand shift (5%, 10%, 15 %) maybe based on some statistics  Future Scope:   Life cycle assessment: looking at VKT, analyzing emissions, vehicle life and many more parameters.  Prof. F. CiARI; P.K. Pandey & R. HASSAN 6 Assessing the Effect of Demand Shift on Fully Shared Autonomous Transportation System

7. Prof. F. CiARI; P.K. Pandey & R. HASSAN 7 Assessing the Effect of Demand Shift on Fully Shared Autonomous Transportation System Day-1 Initialization

8. 8 Thank You Any Questions Assessing the Effect of Demand Shift on Fully Shared Autonomous Transportation System