Remote Driving Handover for Unexpected Autonomous Road Conditions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous driving techniques rely on machine learning models that struggle to handle unexpected or unique driving conditions not covered in their training data, raising safety concerns due to potential inaccuracies in autonomous driving decisions.
Innovation Solution
A system that establishes a remote control link between a vehicle and a remotely operated computing device, allowing real-time driving condition data to be shared and enabling an operator to manually control the vehicle's steering, accelerating, or braking, thereby addressing unexpected driving conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If autonomous driving capabilities are implemented using machine learning models, then the vehicle can operate autonomously, but the system fails to handle unexpected driving conditions not covered in training data
Solution Approach 1:
The patent introduces a remote computing device as an intermediary between the vehicle's autonomous driving system and human operators. This mediator enables seamless transition from autonomous to remote-controlled operation when unexpected conditions are detected, combining automated operation with human oversight to resolve the reliability issue.
Solution Approach 2:
The system dynamically adjusts the level of automation based on driving conditions. It transitions from fully autonomous operation to remote-controlled operation when anomalies are detected, making the automation extent adaptable rather than fixed, thereby maintaining reliability across varying conditions.
2Reliability
If a remote control link is established between the vehicle and computing device, then human operators can handle unexpected conditions, but the system complexity increases
Solution Approach 1:
The remote computing device serves multiple functions: it monitors driving conditions, detects anomalies, establishes communication links, and controls vehicle operations. This multi-functionality reduces the need for separate dedicated components for each function, thereby managing system complexity while maintaining reliability.
Solution Approach 2:
The system automatically detects when remote control is needed and initiates the connection process without requiring manual intervention. The vehicle's control system self-manages the transition to remote operation, reducing operational complexity despite the added remote control capability.
3Productivity
If machine learning models are used for autonomous driving decisions, then the vehicle can make autonomous decisions, but the decisions may be inaccurate in unfamiliar conditions
Solution Approach 1:
The system continuously monitors driving conditions and feeds this information back to the autonomous driving model. When the model's confidence in its decisions drops below a threshold or anomalies are detected, the feedback mechanism triggers a transition to remote-controlled operation, ensuring decision accuracy through human oversight when needed.
Data Source
AI summary
Systems/techniques that facilitate semi-autonomous or pseudo-autonomous driving are provided. In various embodiments, a system onboard a vehicle can discover one or more computing devices that are physically remote from the vehicle but that are within electronic communication range of the vehicle. In various aspects, the system can establish a first remote control link between the vehicle and a first computing device of the one or more computing devices, such that steering, accelerating, or braking of the vehicle are operated autonomously or by a physical driver prior to establishment of the first remote control link, and such that the steering, accelerating, or braking of the vehicle are remotely operated by the first computing device after establishment of the first remote control link.