Scene-Aware Navigation System for Contextual Route Guidance

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing navigation systems fail to provide clear and intuitive route guidance to drivers by not utilizing real-time information about static and dynamic objects in the vehicle's vicinity, leading to confusion and potential safety issues, especially when static objects are obstructed or no longer recognizable.

Innovation Solution

A scene-aware navigation system that processes real-time data from multiple sources, including cameras, microphones, and range detectors, to identify salient static and dynamic objects, and generates context-based driving instructions using machine learning methods, which can be spoken or displayed, and allows for two-way interaction with the driver.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If GPS-based route guidance is used to provide turning instructions with distance and direction, then the driver receives basic navigation information, but the guidance becomes confusing when the driver cannot easily recognize the street name or when multiple streets are in close proximity

Engineering Contradiction:
Improveroute guidance clarityVSAvoiddriver recognition of reference points
Core Design Contradiction:
Loss of informationVSEase of operation

Solution Approach 1:

The system uses camera-captured images as an intermediary between the GPS navigation system and the driver. Instead of relying solely on abstract GPS instructions, the system overlays visual annotations (arrows, highlights) on real-time camera images to bridge the gap between digital navigation data and physical road recognition, making it easier for drivers to identify the correct street among multiple options

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system transitions from two-dimensional map-based GPS instructions to a third-dimensional perspective by using real-time camera views that show the actual physical environment. This adds the dimension of visual realism and spatial context, allowing drivers to recognize reference points in their actual field of view rather than translating abstract map coordinates to physical locations

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Loss of information

If stored map information is used to indicate turning points with points of interest, then the driver receives contextual navigation cues, but the guidance becomes unreliable when trees or vehicles hide the point of interest or when the stored information is outdated

Engineering Contradiction:
Improvenavigation contextVSAvoidvisibility and accuracy of reference points
Core Design Contradiction:
Loss of informationVSReliability

Solution Approach 1:

The system replaces static, pre-stored map information with dynamic real-time camera images. The reference points are no longer fixed database entries but are dynamically captured and displayed in the current scene, ensuring they reflect the actual present state of the environment including any occlusions or changes, thereby maintaining reliability through continuous updates

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system performs preliminary detection and annotation of reference points using the camera before the driver needs to identify them. By pre-processing the visual scene to highlight and annotate relevant objects (buildings, signs, vehicles) in advance, the system ensures that when the driver needs navigation guidance, the reference points are already prepared and visually marked for immediate recognition

Inventive Principle:
Principle #10Preliminary action

3Loss of information

If real-time camera images with overlaid graphical elements are used to indicate the route, then the driver receives visual route guidance, but the driver must take their eyes off the road to see the guidance

Engineering Contradiction:
Improveroute visualizationVSAvoiddriver distraction
Core Design Contradiction:
Loss of informationVSObject-affected harmful factors

Solution Approach 1:

The system positions the camera and display arrangement so that the route guidance information is presented in the driver's natural field of view, creating an equipotential viewing condition where the driver can perceive both the road ahead and the navigation annotations without changing their gaze direction or focal plane, thereby eliminating the need to shift attention between different spatial locations

Inventive Principle:
Principle #12Equipotentiality

4Measurement precision

If multiple sensors and machine learning methods are used to process real-time data and identify salient objects, then the system provides accurate and context-aware navigation, but the device complexity increases

Engineering Contradiction:
Improveobject detection accuracyVSAvoidsensor and processing system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system merges multiple sensor inputs (camera, microphones, range detectors) and processing functions (real-time data processing, machine learning-based object identification, context analysis) into a single integrated navigation system. This consolidation allows the complex components to work together synergistically, where the combined system achieves higher measurement precision than individual components could provide alone, while the integrated architecture manages complexity through unified processing

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP3994426B1Method and system for scene-aware interaction
Publication Date: 2023.05.31 MITSUBISHI ELECTRIC CORP
  • EP3994426B1 patent drawingFigure 1A
  • EP3994426B1 patent drawingFigure 1B
  • EP3994426B1 patent drawingFigure 1C

AI summary

A navigation system configured to provide driving instructions to a driver of a moving vehicle based on real-time description of objects in a scene pertinent to driving the vehicle is provided. The navigation system includes an input interface configured to accept a route for driving the vehicle, a state of the vehicle on the route at a current instance of time, and a dynamic map of a set of salient objects pertinent to the route of the vehicle at the current instance of time, wherein at least one salient object is an object perceived by a measurement system of the vehicle moving on the route between a current location at the current instance of time and a future location at a future instance of time, wherein the set of salient objects includes one or multiple static objects and one or multiple dynamic objects, a processor configured to generate a driving instruction based on a description of a salient object in the dynamic map derived from a driver perspective specified by the state of the vehicle; and an output interface configured to render the driving instruction to the driver of the vehicle.