Abstract: System and techniques for vehicle occupant monitoring are described herein. Sensor data, that includes visual image data, is obtained from a sensor array of the vehicle. An object carried by the vehicle is detected from the visual image data. A safety event for the vehicle may be identified based on the object detection and an operational element of the vehicle is altered in response to detecting the safety event.

Abstract: An exemplary method includes obtaining vehicle data comprising environmental perception data indicating a risk assessment regarding one or more perceived elements of an environment surrounding a vehicle; obtaining driver perception data regarding a driver inside the vehicle; determining an integrated risk assessment based on the vehicle data and the driver perception data; and determining an Operational Design Doman (ODD) compliance assessment of the vehicle at least based on the determined integrated risk assessment.

Abstract: Techniques are disclosed for the exploration of environments for the estimation and detection of hazards or near hazards within the environment and the mitigation of hazards therein. The exploration of the environment and mitigation of hazards therein may use one or more autonomous agents, including a hazard response robot. The estimation of the hazards may use a policy learning engine, and the hazards may be detected, and the associated risks therefrom, may be determined using a hazard estimation system.

Abstract: The automated driving perception systems described herein provide technical solutions for technical problems facing navigation sensors for autonomous vehicle navigation. These systems may be used to combine inputs from multiple navigation sensors to provide a multimodal perception system. These multimodal perception systems may augment raw data within a development framework to improve performance of object detection, classification, tracking, and sensor fusion under varying external conditions, such as adverse weather and light, as well as possible sensor errors or malfunctions like miss-calibration, noise, and dirty or faulty sensors. This augmentation may include injection of noise, occlusions, and misalignments from raw sensor data, and may include ground-truth labeling to match the augmented data. This augmentation provides improved robustness of the trained perception algorithms against calibration, noise, occlusion, and faults that may exist in real-world scenarios.

Abstract: Various systems and methods for detecting risk conditions in a physical workspace. An apparatus can include an interface to receive smart sensor signals from at least one autonomous mobile entity (AME) in the physical workspace. The apparatus can also include processing circuitry coupled to the interface to detect a risk condition associated with the at least one AME, based on the smart sensor signals, relative to a user device associated with a human present in the physical workspace. The processing circuitry can also detect a direction of the risk condition relative to the user device and cause a notification to the first user device. The notification can indicate the direction of the risk condition relative to the user device. Other systems, methods and apparatuses are described.

Abstract: Systems and techniques for controlling a mobile robot. In an example, the system may include a processor and memory, including instructions, which when executed by the processor, cause the processor to determine a state of the mobile robot. The state of the mobile robot may be determined using at least one of: data provided by the mobile robot or data captured by one or more sensors proximate to or attached to the mobile robot. The system may determine a state of an object proximate to the mobile robot using the data captured by the one or more sensors, and identify information relating to one or more available stopping points. The system may identify a condition of the mobile robot that requires the mobile robot to stop and issue a command to the mobile robot to navigate to a particular one of the one or more available stopping points.

Abstract: Disclosed herein are systems, devices, and methods of a system that may adapt sensing capabilities based on the risk of the situation. The system may determine a risk estimation of an object in a work environment of a robot based on sensor data and a measurement uncertainty of a sensing system configured to collect the sensor data, wherein the sensor data is indicative of the object in the work environment. The system may also, based on the risk estimation, select a next operation mode from a plurality of operation modes for the sensing system.

Abstract: Systems and techniques for reliable real-time deployment of robot safety updates are described herein. Condition data may be collected for an environment in which a robot is operating. The robot may be classified with a condition type. Condition type data selected from the condition data may be analyzed based on the condition type to calculate a safety risk level for the robot. A microservice may be identified to provide a robot safety rule for the robot based on the safety risk level. The microservice may be identified using a safety prediction model generated from the condition type data.

Abstract: A computing device, including: a memory configured to store computer-readable instructions; and processing circuitry configured to execute the computer-readable instructions to cause the computing device to: generate an environmental condition model (ECM) based on collected environmental attribute data of a physical environment; deploy the ECM to estimate environmental risk for an autonomous agent based on the environmental attribute data and a position of the autonomous agent within the physical environment; and transmit information related to the estimated environmental risk to the autonomous agent to adapt its operation to reduce operational risk.

Abstract: Disclosed herein are systems, devices, and methods of a safety system for analyzing and improving the safety of environments that may be shared between robots and humans. The safety system may receive a safety envelope that includes a reachable set of locations at an expected position of an object at a prediction time. The receiver may also receive a perception prediction that may be based on the safety envelope and may include environmental information associated with the object at the expected position at the prediction time. The safety system may also include a processor that generates an instruction to move a robot according to a safe movement instruction based on whether a perception check exceeds a predetermined threshold, wherein the perception check may be based on a difference between the perception prediction and sensor information indicative of the environment of the object at the prediction time.

Abstract: Various aspects of techniques, systems, and use cases include robot safety. A device in a network may include processing circuitry and memory including instructions, which when executed by the processing circuitry, cause the processing circuitry to perform operations. The operations may include collecting telemetry data for a robot, the robot operating according to a path control plan generated using reinforcement learning with a safety factor as a reward function, and detecting that a safety event, involving a robot action, has occurred with the robot and an object. The operations may include simulating a recreation of the safety event to determine whether a simulated action matches the robot action.

Abstract: Disclosed herein are systems, devices, and methods of a safety system for analyzing and improving the safety of collaborative environments in which a robot may interact a human. The safety system may determine a monitored attribute of a person within an operating environment of a robot, where the monitored attribute may be based on received sensor information about the person in the operating environment. In addition, the safety system may determine a risk score for the person based on the monitored attribute. The risk score may be defined by (1) a collision probability that the person will cause an interference during a planned operation of the robot and (2) a severity level associated with the interference. The safety system may also generate a mitigating instruction for the robot if the risk score exceeds a threshold level.

Abstract: Devices and methods for determining an action in the presence of road users are provided in this disclosure. A device may include a processor. The processor may be configured to access environment information including an indication of a size of road users intersecting with a predetermined route of a vehicle in a road environment. The processor may further be configured to prioritize an anticipated movement of at least one of the road users over a predicted movement of the vehicle within the predetermined route based on the size of road users. The processor may further be configured to determine a vehicle action allowing the anticipated movement of the at least one road user.

Abstract: According to various aspects, a safety module is described including: one or more processors configured to receive road information representing a geometry of one or more roads in a Cartesian coordinate system, determine a lane coordinate system based on the received road information, the lane coordinate system including a plurality of lane segments arranged along a longitudinal direction and along a lateral direction of the lane coordinate system, wherein a length information and a width information are assigned to each of the lane segments, and determine a potential collision event based on the lane coordinate system.

Abstract: Disclosed herein are systems, devices, and methods of a safety system for monitoring the in-vehicle safety of internal objects within a vehicle. The safety system generates a digital twin of the interior objects from vehicle configuration data indicating a configuration of an interior environment of the vehicle, from interior object data associated with the interior object within the interior environment of the vehicle, and from vehicle situation data that indicates an operating status of the vehicle. The digital twin is an abstract model of the interior objects within the interior environment of the vehicle, and the safety system generates, based on the digital twin, a safety score associated with an operating behavior of the vehicle. Based on the safety score, the safety system determines a target level for the operating behavior.

Abstract: Techniques are disclosed to facilitate path planning safety guiding robots (SGR) for safely navigating and guiding humans through autonomous environments having other autonomous agents such as stationary and/or mobile robots.

Abstract: Disclosed herein is an active cyclist and/or pedestrian safety system for identifying, warning, and reacting to dangerous situations that may be experienced by cyclists and/or pedestrians. The safety system may receive first information indicative of a head position of a cyclist and determine, based on the first information, a field of view of the cyclist. The safety system may receive second information indicative of an environment of the cyclist and determine, based on the field of view and the environment, an expected trajectory of the cyclist in a next road segment. The safety system may also determine, based on the field of view and the environment, a risk probability for the expected trajectory of the cyclist in the next road segment. The safety system may also generate an instruction to transmit a warning to the cyclist if the risk probability exceeds a threshold value.

Abstract: A monitoring system may include a memory having computer-readable instructions stored thereon and a processor operatively coupled to the memory. The processor may read and execute the computer-readable instructions to perform or control performance of operations. The operations may include receive, prior to a collision involving a vehicle, sensor data representative of a feature of an internal environment and determine the collision has occurred. The operations may include automatically instruct, based on the collision, a sensor to generate another sensor data representative of another feature of the internal environment. The operations may include receive the another sensor data from the sensor and compare the sensor data and the another sensor data to accident data corresponding to previous accidents. The accident data may include a diagnosed injury and an accident severity of each of the previous accidents. The operations may include determine a severity of the collision based on the comparison.

Abstract: Disclosed herein are systems and methods for detecting and mitigating inappropriate behavior. The systems and methods may include receiving data. Using the data a harassment score and/or classification for a behavior may be determined. Using the harassment score and/or classification, a determination may be made as to when the harassment score and/or classification for the behavior exceeds a threshold. When the threshold is exceeded, a protection system and/or action engine may be activated to mitigate the inappropriate behavior.

Abstract: Disclosed herein is a vehicle handover system that monitors an environment of a vehicle. The vehicle handover system receives a transition request to change control of the vehicle from an automated driving mode to a passenger of the vehicle. The vehicle handover system detects a key event that may be relevant to the transition request and the detection of the key event is based on the monitored environment. The vehicle handover system may generate a handover scene that includes images associated with the key event, and the images include an image sequence over a time-period of the key event. Before the vehicle handover system changes control of the vehicle from the automated driving mode to the passenger, the handover scene is displayed to the passenger.