Autonomous Vehicle Communication Lighting

Assisting autonomous driving vehicles by leveraging remote assistance when the vehicle's onboard systems cannot determine optimal strategies for dealing with certain traffic factors. The vehicle identifies traffic factors impacting driving using its sensors, and if it can't determine a response, it sends the factor info to a remote device. The remote device analyzes the factors and provides guidance back to the vehicle. This allows the vehicle to overcome sensing limitations and improve safety by leveraging off-board intelligence.

Source

Notifying information to autonomous vehicles without relying on mobile network connectivity. The system uses sensors around intersections to monitor traffic conditions. Based on those conditions, it generates driving instructions for vehicles entering the intersection. These instructions are displayed as code information on signs around the intersection. The autonomous vehicles can capture the codes and decode them to obtain the instructions. This allows providing localized, real-time guidance to autonomous vehicles without requiring mobile connectivity.

Source

Outputting warning signals from autonomous vehicles to road users that consider gestures and vocalizations from the vehicle occupant and the road user's viewing direction. The method involves capturing gestures and vocalizations from the vehicle occupant, detecting the road user, determining the road user's viewing direction, and outputting a warning signal based on the captured gestures/vocalizations and road user viewing direction. This prevents misunderstandings where a driver's gesture or comment is interpreted as a control command by the road user. The warning signals inform the road user that the vehicle is autonomous and the driver doesn't have manual control. The method can also factor in environmental objects and priority rules to further refine the warning signals.

Source

Enabling communication between autonomous vehicles and nearby pedestrians using brain-computer interfaces (BCIs) to improve vehicle navigation and safety around pedestrians. Pedestrians wear BCIs that analyze brainwave signals and broadcast them to nearby vehicles. Vehicles receive the signals, predict pedestrian movements, and adjust driving actions accordingly. Pedestrians can also see vehicle responses on AR devices. This allows vehicles to anticipate pedestrian actions and respond proactively.

Source

Automated vehicle system that improves safety in situations where the ego vehicle does not have right of way. The system uses onboard sensors to analyze actions of nearby vehicles and calculate a confidence level for whether they are actually giving way. This confidence is then used to guide the ego vehicle. Actions like yielding lane position, flashing lights, and slowing down are analyzed. By considering the whole context, the system can distinguish genuine yielding intentions from false positives. This avoids relying solely on V2V communications that may be incorrect or delayed.

Source

Predicting environmental conditions like construction, traffic signals malfunction, stalled vehicles, etc. for an autonomous vehicle based on the behavior of other road actors. The system detects the current behavior of a road actor like a car, pedestrian, or cyclist in response to an environmental state. By analyzing the behavior, it infers the environmental state and plans driving policy for the autonomous vehicle accordingly. This allows the vehicle to anticipate and adapt to road conditions that may be outside its direct perception.

Source

Determining the context and intent of a remote vehicle when no connected vehicle technology is available. The method involves using cooperative infrastructure sensing messages to gather vehicle parameters like position, speed, etc. Associating the vehicle with the lane it's driving on based on map data. Calculating possible maneuvers, exits, and speed limits for that lane. Then using the vehicle parameters, lane info, and speed limits to determine the context (short history) and intent (short prediction) of the remote vehicle.

Source

A system for detecting signals from other vehicles and pedestrians, determining appropriate responses, sending responsive signals, and reacting to received responses. This enables autonomous vehicles to respond to gestures, flashing lights, etc., from other road users indicating their intentions and also to initiate signaling to indicate their own intentions to others. It involves detecting signals, determining intended actions, sending response signals, and taking actions based on received responses.

Source

An adaptive illumination system for autonomous vehicles automatically adjusts the headlights to optimize visibility in low-light conditions. The system uses sensors to detect road conditions and determine if specific areas should be illuminated. If areas are deemed insufficiently lit, the headlights are adjusted to illuminate those areas. This adaptive illumination helps improve safety and visibility in dark environments, turns, tunnels, and bridges where sensors may not provide full coverage.

Source

Pedestrian intention prediction method for vehicles using inter-vehicle collaborative computing to improve safety by sharing predictions between vehicles. A core vehicle requests nearby vehicles with similar heading to send computing resources. The core vehicle then distributes pedestrian detection, attitude estimation, and environment extraction tasks to surrounding vehicles. The core vehicle combines their results to predict pedestrian intentions. It then shares intention and vehicle decision info with rear vehicles to allow them to react in advance.

Source

Predicting the behavior of nearby vehicles to improve autonomous driving decision making. The system receives sensor data from the host vehicle, extracts features to detect nearby vehicles, generates trajectories, and uses a trained model to predict intentions. These predictions are output to other subsystems like motion planning. The goal is to enable the autonomous vehicle to anticipate nearby vehicle actions and respond accordingly.

Source

Autonomous vehicle system that interprets traffic signals and moves through intersections without human input. The system identifies the lane occupied by the vehicle and associates a semantic meaning to the traffic signal based on the lane attribute. It then determines the appropriate maneuver for the vehicle based on the semantic meaning. This allows the vehicle to accurately interpret and respond to traffic signals in intersections, even if the signal placement is not optimal.

Source

Driver assistance system and method to help a driver take over an automated vehicle by matching their perception of the driving environment to the vehicle's sensors. The system captures vehicle and traffic data during autonomous driving and presents it to the driver in multimodal formats like visual, audio, and checklists. It compares the driver's responses to the transmitted information to assess readiness for manual driving. If matching is good, the system allows handover. The method involves acquiring vehicle and traffic data, transmitting it to the driver, and comparing perceived vs transmitted info to gauge readiness.

Source

Communication system for autonomous vehicles to safely interact with road users like pedestrians and cyclists. The system uses visual and auditory signals from the vehicle to warn and guide nearby people. It adapts the signals based on risk assessments of identified road users. The signals are targeted, customized, and escalated to draw attention. The vehicle's environment perception system detects road users, determines scenarios, and calculates risk ratings. An embedded system processes the data to adapt and output signals via lights, displays, speakers, etc. This replaces direct human-to-human interaction lost in autonomous driving.

Source

Determining whether an autonomous vehicle should yield to pedestrians based on their intent to cross the road. The system uses predictions of pedestrian crossing intentions alongside nominal crossing predictions to determine if the vehicle should yield before reaching the crossing point. If the pedestrian intends to cross and the nominal prediction indicates they will, the vehicle should yield to let them pass. But if the pedestrian intends to cross and the nominal prediction is uncertain, or if the pedestrian doesn't intend to cross, the vehicle shouldn't yield. This avoids unnecessary yielding that could obstruct traffic or create hazards.

Source

Automatic vehicle function control using optical traffic signals to enhance safety and convenience. The method involves using AI to recognize and interpret various optical traffic signals like warning triangles or flashing lights. The system detects these signals from other vehicles or infrastructure using sensors, and the AI processes them to determine the traffic situation. This information is then used to automatically control vehicle functions like acceleration and braking to respond appropriately to the situation. The AI can also verify the signals against alternative sources like camera-based traffic sign recognition.

Source

System for autonomous vehicles to communicate with pedestrians and cyclists to improve safety and trust. The system uses onboard sensors, computers, and displays to provide intuitive and unambiguous messages to vulnerable road users. It shows avatars of people crossing in real-time on the vehicle's exterior displays. This lets pedestrians see that the autonomous vehicle recognizes and is considering them. The displays also indicate if the vehicle is starting or stopping at a crossing. This enhances transparency and trust between the vehicle and vulnerable road users.

Source

Enabling autonomous vehicles and vehicles with advanced driver assistance systems (ADAS) to understand and respond to non-verbal human communication, like gestures and facial expressions. The system detects situations where human-human communication is likely, such as crossing the road or changing lanes, and monitors the vehicle's surroundings for non-verbal cues from pedestrians. It can then respond appropriately, like stopping or yielding, to acknowledge and accommodate the human's intent. The system learns these interactions from a dataset of recorded human-human scenarios.

Source

Directing a vehicle occupant's attention during automated driving to critical areas that may not be fully detected by the vehicle sensors. The system determines specific areas in the vehicle environment that require attention, like blind spots or upcoming objects, and checks if the occupant is looking there. If not, it signals the occupant to look by displaying warnings near those areas. This improves safety in automated driving by ensuring occupants actively monitor critical areas.

Source

Wireless communication system for vehicles that allows predictive interactions between vehicles based on sensor data. The system enables vehicles to share intent and negotiate maneuvers using V2X communication before actions are taken. This allows early warning and coordination to prevent collisions. The system uses sensor fusion and machine learning to infer intent like exiting a parking space or opening a door before initiating the action. Vehicles can then proactively share this intent via V2X to nearby vehicles and pedestrians. If another vehicle detects a safety issue, it can request the vehicle to stop the action.

Source