Join GitHub today
GitHub is home to over 28 million developers working together to host and review code, manage projects, and build software together.Sign up
Advantages of Round-Robin eBike/eScooter space allocation, with block maximums #4
As an alternative to city-wide quotas, round-robin space assignment would make more vehicles available to underserved areas. Using per-block maximums will ensure that vehicles are more evenly distributed throughout the city.
Vehicles with higher user ratings will be given a weighted preference in the round-robin selection process. This will encourage investment in quality vehicles matching consumer preferences.
Block maximums could be 10 vehicles per type (bike, e-bike, e-scooter, etc) per square block, with racks or geofenced areas storing up to 5 vehicles. Allocated parking areas could be approximately 1/4 of the block from each corner, and/or in the center of the block.
Preset maximums could be based on rider history and population density. Allocation percentages may be defined for the four sides of each block, so vehicles are not all placed on one side. Block maximums could be lowered until vehicles are used at least 3 times a day, or at least 2/3 of vehicles are used daily.
Round-Robin with Block Maximum vs Citywide Quotas
The NACTO-Shared-Active-Transportation-Guidelines states "In order to ensure that Shared Active Transportation Companies provide a reliable, convenient transportation option for citizens, cities should consider how many small vehicles should be made available."
However, setting citywide quotas becomes unnecessary with round-robin allocation of parking and per-block maximums.
Weighted Round-Robin - How it Works
Riders would indicate vehicle model preferences based on a scale of 0 to 100%. The default value would be 50%. Vehicles rated at 100% would be their top choices.
Riders would also indicate where they would be likely to start riding the vehicle.
Each evening, blocks would be assigned for the following day for all operators. First a block would be randomly selected. Next an operator would be assigned based on a weighted average.
Each operator would start with 1 ticket in the lottery for the block for each of their vehicles that remain unassigned.
The preference votes of riders within one block of the space would increase or decrease the weight of 1 ticket per vehicle.
If all riders in the city gave a vehicle model a 100% vote (within the previous week), the operator would have 2 tickets in the lottery for the block (per vehicle) using the following weighted formula:
1 of bike model A @ 100% (Everyone loves it within 1 block) x 2 = 2 entries in lottery for block
To keep preferences fresh, the number of weeks since the rider's vote could reduce the impact of the rating, pulling it closer to the mean of 50%. A vote older than three years could be considered neutral.
Operators would have access to stats for each side of a block - to position vehicles based on preference. Portions of the data would be open, with restrictions to protect privacy.
Operators could occasionally exceed a block maximum by paying a small fee to the city.