Independent Study: Evan

Macroscopic route planning, bicycle parking placement, and resource allocation planning

Research Objective

Users demand for rapid arrival of KanGo and rapid delivery of their bicycles.
Business entity deploying KanGo demand to lower deployment cost, both in terms of number of KanGos and number of parking spots.

The indepedendent study aims to take these two pulling forces into account, with the context of a pre-mapped area, to check for the validity of deploying KanGo in real life

Insights

- Dijkstra’s algorithm finds shortest path and its path length between two points in the pre-mapped area extremely fast: Estimated <1ms per pair
- Closed-form solution to find the optimal placement of N parking spots using existing algorithms used for graph splitting
- Expected waiting time metric for each placement of parking spots as generated above, taking nto account waiting time increase factors such as peak hours and mismatch of park/retrieval location
- Interactive two-way calculator for KanGo deploying party to (a) optimize for lowest waiting time given their budget (b) estimate how much budget is required to meet a certain waiting time target

Graph-based approach

Best: 8W, followed by 10W 3.73 min avg. waiting time

Pixel-based approach

Best: 8W, followed by 5W 19.33 avg. steps (KanGo estimated to take 5 step per minutes)

Interactive calculator

Significance of Research

KanGo can potentially work in real life in B2C or B2B2C configuration
VEEEEE! may choose to establish its own company and operate in B2C mode. The findings enables cost transparency and proves working possibility to its investors
VEEEEE! may choose to work with other business entities (such as property management) which wish to enhance active mobility compatibility in their region. The findings eases negotiation and aligns expectation between both parties.