Vehicle Map Adaptation for HD and Non-HD Autonomous Driving
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in operating efficiently with high-definition (HD) maps due to their unavailability or outdatedness in certain geographical areas, and maintaining both HD and non-HD map-based computing systems is costly and complex.
Innovation Solution
A vehicle computing system is retrofitted to adapt to multiple map types by leveraging sensor data to generate synthetic high-definition maps, using machine learning models like generative adversarial networks (GANs) to enhance lower-definition maps in real-time, allowing a single processing system to operate with both HD and non-HD maps.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a vehicle computing system is designed to operate with high-definition maps, then driving decision accuracy is improved, but system complexity and cost increase when needing to support multiple map types
Solution Approach 1:
The computing system is designed with a universal processing architecture that can handle both HD and non-HD map data through the same perception, prediction, planning, and control stacks. A map adaptation module converts non-HD map data into a format compatible with the HD-map-dependent processing system, allowing single-system operation across multiple map types without duplicating computing infrastructure.
Solution Approach 2:
A map adaptation module serves as an intermediary component between the input map data (whatever type) and the downstream processing stacks. This module transforms non-HD map data into enhanced representations that are compatible with the HD-map-dependent processing architecture, enabling seamless operation without requiring separate computing stacks for different map types.
2Adaptability or versatility
If both HD and non-HD map-based computing systems are maintained, then operational versatility is improved, but cost and system complexity increase
Solution Approach 1:
The system employs a universal processing architecture where a single computing stack handles both HD and non-HD map operations. The map adaptation module enables this universality by transforming non-HD map data into a format that the existing HD-map-dependent processing stacks can consume, eliminating the need for maintaining separate computing systems for different map types.
Solution Approach 2:
The patent merges the functionality of multiple map-type-specific computing stacks into a single unified system. By combining the map adaptation capability with the existing processing stacks, the system consolidates what would otherwise require separate HD-map and non-HD-map processing infrastructure into one integrated computing platform.
3Reliability
If HD maps are used in areas where they are unavailable or outdated, then driving safety is compromised, but using alternative map types reduces decision accuracy
Solution Approach 1:
The system dynamically adapts to the availability and quality of map data in different geographical areas. When HD maps are unavailable or outdated, the map adaptation module dynamically transforms available non-HD map data into enhanced representations, allowing the processing stacks to operate effectively regardless of the underlying map quality. This dynamic adaptation ensures continuous safe operation without being constrained by map availability.
Data Source
AI summary
Systems and methods for retrofitting a vehicle computing system dependent on a certain type of maps to operate with other type(s) of maps are provided. For example, a method performed by a vehicle may include receiving, from one or more sensors of the vehicle, first sensor data associated with a surrounding environment of the vehicle; receiving first map data of a first map type; and retrofitting a vehicle controller of the vehicle that is based on a second map type different from the first map type to operate on the first map data, where the retrofitting includes adapting the first map data using the first sensor data to generate second map data associated with the second map type; and determining, by the vehicle controller, an action for the vehicle based at least in part on the generated second map data.