Map Shard Route Precomputation for Memory-Limited Autonomous Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in accessing and processing highly complex map information, which requires large amounts of RAM, making it impossible to retrieve all map information when routing between distant locations, such as different states, due to computational intensity and memory constraints.
Innovation Solution
The method involves partitioning the map into smaller 'map shards' that can be loaded and unloaded as needed, with pre-computation of optimal routes across individual shards, allowing vehicles to determine routes using only the necessary shards and associated costs, reducing computational and memory requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all map information is retrieved for routing between distant locations, then routing accuracy is improved, but memory requirements become impossible to satisfy
Solution Approach 1:
The map is divided into multiple shards, each representing a geographic region. Instead of loading the entire map into memory, only relevant shards containing the start location, end location, and intermediate regions are loaded and processed independently, then merged to form the complete route.
Solution Approach 2:
The routing problem is transformed from a global search across the entire map to a multi-dimensional approach where routes are computed separately for each shard (local dimension) and then integrated (global dimension), reducing memory requirements while maintaining routing accuracy.
2Speed
If pre-computation is performed for all possible routes, then routing speed is improved, but computational complexity becomes unmanageable
Solution Approach 1:
Route computation is segmented into independent shard-level pre-computations. Each shard pre-computes routes between its own port nodes (entry/exit points), and these local route sets are stored. During actual routing, pre-computed shard routes are combined rather than computing the entire route from scratch, significantly reducing computational complexity.
Solution Approach 2:
Routes within each shard are pre-computed and stored in advance, including the costs associated with traversing between port nodes. This preliminary computation eliminates the need for expensive real-time pathfinding across the entire map, improving routing speed while keeping per-shard computational complexity manageable.
3Manufacturing precision
If detailed map information is processed in real-time, then route optimization is improved, but energy consumption increases
Solution Approach 1:
Route optimization data is pre-computed for each shard including cost information for traversing between port nodes. This eliminates the need for real-time optimization calculations during vehicle operation, significantly reducing energy consumption while maintaining route optimization quality.
Solution Approach 2:
Instead of processing the entire map in real-time, only partial map data (relevant shards) is processed and pre-stored. The pre-computation performs slightly more work locally within each shard, but this partial action approach dramatically reduces the total energy required compared to full-map real-time processing.
Data Source
AI summary
Aspects of the disclosure relate to pre-computing routes for autonomous vehicles using map shards. For example, a shard from a plurality of shards of a map may be selected. Each shard including a plurality of nodes and edges connecting pairs of nodes of the plurality of nodes, and each node of the plurality represents a location. A plurality of port nodes for the shard are identified. Each port node has an edge that enters into the selected shard or exists the selected shard. For each port node of the plurality having an edge that enters into the selected shard, optimal routes to each other port node of the plurality having an edge that exits the selected shard may be determined. The optimal routes for the selected shard may be sent to the autonomous vehicles in order to enable the autonomous vehicles to use the optimal routes to determine routes.