Input
In our case, we modeled autonomous bicycles, the used Boston-Cambridge as the scenario for our simulation, and the trip data was created based on real BlueBikes trips. You can customize any of these:
GIS data#
You will need GIS data of your chosen city/area:
- A shapefile containing the buildings: it can be downloaded using the OSMbuildings service.
- A shapefile containing the roads: it can be downloaded using the Overpass API.
| Buildings | Road network |
|---|---|
 |  |
User demand#
The user demand data must be in a .csv format that has the following colums for each user:
- start_lat : latitude at the start location, e.g. 42.36114.
- start_lon : longitude at the start location, e.g. -71.05703.
- target_lat : latitude at the target location, e.g. 42.37003.
- target_lon : longitude at the target location, e.g. -71.04555.
- start_time : trip departure time in seconds, relative to the simulation start, e.g. 3600 means it starts 1 hour from the simulation start.
This .csv must be saved under the data folder. The name of the .csv must be specified in main.py:
users_path = os.path.join("data", "user_trips.csv")The demand considered for the simulation is based on Bluebikes public bike-sharing system usage data. The user generation process takes advantage of this historical usage data and the buildings' spatial data. Buildings data is used to generate users' origin and destination locations inside the buildings, scattering users in buildings 300 m around stations. This process can be found UserGeneration.py under the folder Preprocessing. UserGeneration.py is not called every time the simulation runs because it is a time consuming process, but it can be run manually to create a new user demand file.
Stations#
The station data must be in a .csv format that has the following colums for each station:
- "Longitude"
- "Latitude"
- "Total docks"
The location of docking and charging stations for this simulation has also been extracted from Bluebikes. Similarly to the user demand data, the .csv containing the information related to the stations must be saved under the data folder. The name of the .csv must be specified in main.py:
stations_path = os.path.join("data", "bluebikes_stations_07_2020.csv")This file is used in BikeGeneration.py under the folder Preprocessing. This code is automatically called for each simulation. In the three systems bikes are generated randomly at stations, with a probability proportional to their capacity.
Note
In the station-based system, the total number of bikes is limited to be smaller than the total number of docks in the system. In the dockless and the autonomous systems, bikes are initially located next to stations, in a quantity proportional to the number of docks available at each station, and considering the fleet size considered in that simulation.
Note
In the dockless system, stations data is only used to locate the bikes at the beginning of the simulation. A different function could be used (random, based on demand, etc.) to generate the initial locations. For the moment, stations data is considered.
Configuration#
Config.js is the file where you will set the configuration of the simulations. This structure contains all the config from the three systems (SB= Station-based, DL= Dockless, AUT= Autonomous):| Parameter | Description | Units | Type of system |
|---|---|---|---|
| "MODE" | 0=Station-based, 1=Dockless, 2= Autonomous | [-] | SB, DL, AUT |
| "NUM_BIKES" | Number of bikes in the system, fleet size | [-] | SB, DL, AUT |
| "WALK_RADIUS" | Maximum distance that a user is willing to walk | [m] | SB, DL |
| "AUTONOMOUS_RADIUS" | Maximum distance that an autonomous bike will do to pick up a user | [m] | AUT |
| "RIDING_SPEED" | Average bike riding speed of users | [km/h] | SB, DL, AUT |
| "WALKING_SPEED" | Walking speed of users | [km/h] | SB, DL |
| "AUTONOMOUS_SPEED" | Average speed of the bike in autonomous mode | [km/h] | AUT |
| "BATTERY_MIN_LEVEL" | Level at which the autonomous bikes go to a charging station | [%] | AUT |
| "BATTERY_AUTONOMY" | Autonomy of the autonomous bikes | [km] | AUT |
| "BATTERY_CHARGE_TIME" | Time that it takes to charge a battery from 0 to 100% | [h] | AUT |
| "INSTANT_BETA" | Probability of a user getting an instant rebalancing; it reflects the amount of rebalancing | [0-1] | SB |
| "INSTANT_MIN_BIKES" | Minimum number of bikes that a station should have for the rebalancing action to remove a bike from that station | [-] | SB |
| "INSTANT_MIN_DOCKS" | Minimum number of docks that a station should have for the rebalancing action to insert a bike in that station | [-] | SB |