Continuous Optimization with Large Neighborhood Search
This article explains Large Neighborhood Search, how we applied it to our ride-sharing service, and an overview of the system architecture.
#optimization
7 posts
How We Automated Dispatch Assignment Work (6 Hours per Day)
As an introduction to our automated dispatch assignment system, this article covers the system overview, assignment logic, monitoring approach, and our plan to introduce deep reinforcement learning.
Order Processing for Ride-Sharing Dispatch
In our service, we accumulate advance reservations and, each time a new order arrives, compute the best ride-sharing combination so dispatch decisions can be made incrementally. This article looks at the kinds of patterns that emerge in this process.
Ride Dispatch Orchestration
In this article, we look at the analogy between Kubernetes and dispatch routing.
On the Smooth Relationship Between Taxis and Buses
In this article, I explain the main parameters used in ridesharing shuttles, show that one extreme becomes taxi-like while the other becomes bus-like, and discuss how we can connect the two smoothly.
State Management in Dispatch Routing
In our pre-booking airport shuttle ridesharing service, we perform routing sequentially as orders come in, and there is a "routing state" that changes over time across all orders. In this article, we will explain the state management of dispatch routing with architecture.
Multi-tenancy in Our Dispatch Service
Multi-tenancy, as seen in many SaaS services, means enabling multiple companies to use a service independently within a single architecture. Combined with the unique nature of dispatch operations, this feature has played an important role in our business expansion.