Cloud-Assisted ADV Detour Control for Dynamic Road Conditions

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing remote control systems for autonomous driving vehicles lack the ability to generate control instructions that are highly safe and reliable, particularly in dynamic driving environments, as they often rely on pre-generated detour routes that may not accurately reflect the current conditions.

Innovation Solution

The system allows the autonomous vehicle to send an assistance request to a cloud server when encountering an impassable road section, which provides a reference detour route, and the vehicle generates specific control instructions based on this route and its current driving environment, ensuring safety and reliability by continuously monitoring and adapting to environmental changes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If the cloud server provides pre-generated detour routes, then the remote control capability is established, but the control instructions may not accurately reflect current driving conditions, reducing safety and reliability

Engineering Contradiction:
Improveremote control capabilityVSAvoidsafety and reliability of control instructions
Core Design Contradiction:
Extent of automationVSReliability

Solution Approach 1:

The system transitions from static pre-generated detour routes to dynamic real-time route generation. The cloud server receives current driving environment data from the autonomous vehicle and generates detour routes that adapt to real-time conditions, ensuring the control instructions reflect the actual driving situation and improve safety and reliability

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system implements a feedback loop where the autonomous vehicle continuously transmits current driving environment data to the cloud server. The server uses this feedback information to generate updated detour routes, ensuring the control instructions are based on the latest environmental conditions rather than outdated pre-generated routes

Inventive Principle:
Principle #23Feedback

2Reliability

If the vehicle generates control instructions based on real-time environment and reference detour route, then the safety and reliability improve, but the system complexity increases

Engineering Contradiction:
Improvesafety and reliability of control instructionsVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system divides the control instruction generation process into two independent modules: the cloud server generates reference detour routes based on environmental data, and the autonomous vehicle's controller generates specific control instructions by combining the reference routes with real-time sensor data. This segmentation allows each module to focus on specific tasks, improving reliability while distributing system complexity across multiple components rather than concentrating it in one complex system

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3968114B1Methods for remote control of an autonomous driving vehicle and devices to support the methods
Publication Date: 2024.01.31 BEIJING BAIDU NETCOM SCI & TECH CO LTD
  • EP3968114B1 patent drawingFigure 1A~1B
  • EP3968114B1 patent drawingFigure 2~3
  • EP3968114B1 patent drawingFigure 4~5

AI summary

A method for remote control of an autonomous driving vehicle (ADV) includes: - sending an assistance request to a cloud server in response to detecting that a current road section in front of the ADV is unable to be passed - obtaining a reference detour route returned from the cloud server - generating control instructions based on the reference detour route and a current driving environment of the ADV - controlling the ADV based on the control instructions.