Ride Booking Allocation via Dynamic ETA-Based Batching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing logistics and transport systems face inefficiencies due to inaccurate pre-zoning of geographic territories, leading to computational complexity, latency, racing conditions, and increased greenhouse emissions, among other issues.
Innovation Solution
Implementing an online clustering approach that dynamically generates non-connected batches using a modified Depth-first search (DFS) to allocate bookings and vehicles, avoiding static sharding and considering real-time data for optimal allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If the geographic territory is divided into smaller zones for batching, then computational complexity is reduced and latency is reduced, but the overall efficiency of optimization deteriorates and racing conditions occur
Solution Approach 1:
The patent implements dynamic batching that adapts to real-time system state rather than using static pre-defined zones. The batching strategy adjusts based on current load, driver availability, and geographic distribution, allowing the system to optimize between computational efficiency and allocation quality dynamically. This resolves the contradiction by making the batching granularity flexible rather than fixed.
Solution Approach 2:
The system changes key parameters such as batch size, batching depth, and geographic boundaries based on real-time conditions. Instead of fixed zones, the patent dynamically adjusts batching parameters to balance computational complexity with optimization efficiency, preventing racing conditions while maintaining low latency through adaptive parameter tuning.
2Measurement precision
If periodic pre-zoning based on historical transaction density is used, then the geographic territory is divided into zones with roughly the same number of expected transactions, but significant computational requirements are introduced and inefficiencies occur when real time data is inconsistent with historical data
Solution Approach 1:
The system uses real-time data to automatically adjust batching decisions without relying on pre-computed historical zones. The batching mechanism serves itself by adapting to current transaction patterns, driver locations, and system state, eliminating the need for separate historical analysis computations while maintaining accurate transaction distribution.
Solution Approach 2:
Instead of performing heavy computational analysis of historical data in advance to create static zones, the patent performs lightweight real-time assessments that naturally account for transaction density and driver availability. The system prepares batches dynamically based on current state, avoiding the computational overhead of historical analysis while achieving better adaptability.
3Device complexity
If static sharding is used to divide bookings into zones, then computational complexity is reduced, but racing conditions occur where the same driver is assigned to multiple bookings
Solution Approach 1:
The patent implements a feedback mechanism where the system tracks driver assignments and batch processing status in real-time. When a driver is assigned to a booking in one batch, this information feeds back to subsequent batch processing decisions, preventing the same driver from being assigned to multiple concurrent bookings. This feedback loop maintains allocation accuracy without requiring complex static zoning.
Solution Approach 2:
The system uses dynamic batch creation that considers current driver availability and assignment status rather than relying on fixed static zones. Batches are formed and processed dynamically with awareness of ongoing assignments, eliminating racing conditions while keeping computational complexity manageable through adaptive rather than rigid zoning.
Data Source
AI summary
A system 100 comprising: a communication server 102; at least one user communication device 104 having an associated user and configured to initiate a booking which includes a booking pickup location; at least one driver communication device 106 having an associated driver vehicle and configured to provide driver vehicle location data; and communication network equipment 108 configured to establish communication with the communications server 102, the at least one user communication device 104, and the at least one driver communication device 106; wherein the communications server 102 comprises at least one processor(s) 116, at least one memory 118, the server 102 being configured, under control of one or more of the at least one processor(s) 116, to execute instructions stored in one or more of the at least one memory 118 to: cluster the proposed booking locations and driver vehicle locations into batches according to an estimated time of arrival between each vehicle and adjacent proposed bookings; cluster a respective batch into further batches if the respective batch has a number of bookings over a predetermined threshold, and allocate the proposed bookings and the available vehicles within each batch separately. A method, a server and various devices are also disclosed.