Systems and methods for an autonomous vehicle are provided. In one aspect, an autonomous vehicle includes a perception sensor and a processor configured to: receive detected roadway conditions data including roadway grade data from the perception sensor, retrieve mapped data having grade data, and determine that the roadway has a grade based on the detected roadway grade data and the retrieved roadway grade data. The processor can be further configured to, in response to determining that the roadway has a grade, determine that the grade of the roadway is greater than or equal to a predetermined high grade value and less than a predetermined grade limit, and in response to determining that the grade of the roadway is greater than or equal to the predetermined high grade value and less than the predetermined grade limit, operate the autonomous vehicle to change lane to a right-most lane.

A passenger may be rather vulnerable to safety risks during pickup and/or drop-off of a passenger by a vehicle. To mitigate or eliminate such risk, the vehicle may determine an endpoint for a vehicle route to pickup or drop-off a passenger at a location. The vehicle may determine an estimated path between the endpoint and the location and may determine a safety confidence score by a machine-learned model for the estimated path and/or may predict a trajectory of a detected object to ascertain whether the estimated path is safe. The vehicle may execute any of a number of different mitigation actions to reduce or eliminate a safety risk if one is detected.

A system 100 for autonomous vehicle operation can include: a low-level safety platform 130; and can optionally include and/or interface with any or all of: an autonomous agent 102, a sensor system, a computing system 120, a vehicle communication network 140, a vehicle control system 150, and/or any suitable components. The system functions to facilitate fallback planning and/or execution at the autonomous agent. Additionally or alternatively, the system can function to transition the autonomous agent between a primary (autonomous) operation mode and a fallback operation mode.

A method may include obtaining one or more inputs in which each of the inputs describes at least one of: a state of an autonomous vehicle (AV) or a state of an object; and identifying a prediction context of the AV based on the inputs. The method may also include determining a relevancy of each object of a plurality of objects to the AV in relation to the prediction context; and outputting a set of relevant objects based on the relevancy determination for each of the plurality of objects. Another method may include obtaining a set of objects designated as relevant to operation of an AV; selecting a trajectory prediction approach for a given object based on context of the AV and characteristics of the given object; predicting a trajectory of the given object using the selected trajectory prediction approach; and outputting the given object and the predicted trajectory.

A crosswalk navigation system for operating an automated vehicle in an intersection includes an intersection-detector, a pedestrian-detector, and a controller. The intersection-detector is suitable for use on a host-vehicle. The intersection-detector is used to determine when the host-vehicle is proximate to an intersection and determine when the intersection includes a cross-walk. The pedestrian-detector is suitable for use on the host-vehicle. The pedestrian-detector is used to determine a motion-vector of a pedestrian relative to the cross-walk. The controller is in communication with the intersection-detector and the pedestrian-detector. The controller is configured to determine a travel-path of the host-vehicle through the intersection, determine when the pedestrian will pass through an intersect-location where the travel-path intersects the cross-walk based on the motion-vector, and operate the host-vehicle to enter the intersection before the pedestrian passes through the intersect-location and to arrive at the intersect-location after the pedestrian passes through the intersect-location.

An autonomous system may selectively displace human driver control of a host vehicle. The system may receive an image representative of an environment of the host vehicle and detect an obstacle in the environment of the host vehicle based on analysis of the image. The system may monitor a driver input to a throttle, brake, and/or steering control associated with the host vehicle. The system may determine whether the driver input would result in the host vehicle navigating within a proximity buffer relative to the obstacle. If the driver input would not result in the host vehicle navigating within the proximity buffer, the system may allow the driver input to cause a corresponding change in one or more host vehicle motion control systems. If the driver input would result in the host vehicle navigating within the proximity buffer, the system may prevent the driver input from causing the corresponding change.

A device and method predict a behavior of a road user. The device is configured to provide at least one hypothesis for the behavior of the road user, to provide, for each hypothesis, a hidden Markov model, the hidden Markov model including, for the particular hypothesis, two hidden states, with one of these hidden states representing the road user following the hypothesis and the other of these states representing the road user not following the hypothesis, and possible observations of the hidden Markov model characterizing, for the particular hypothesis, at least one feature of the road user, and to predict the behavior of the road user depending on the hidden states of the hidden Markov model for the at least one hypothesis.

A system 100 for autonomous vehicle operation can include: a low-level safety platform 130; and can optionally include and/or interface with any or all of: an autonomous agent 102, a sensor system, a computing system 120, a vehicle communication network 140, a vehicle control system 150, and/or any suitable components. The system functions to facilitate fallback planning and/or execution at the autonomous agent. Additionally or alternatively, the system can function to transition the autonomous agent between a primary (autonomous) operation mode and a fallback operation mode.

Systems, methods and apparatus may be configured to implement automatic semantic classification of a detected object(s) disposed in a region of an environment external to an autonomous vehicle. The automatic semantic classification may include analyzing over a time period, patterns in a predicted behavior of the detected object(s) to infer a semantic classification of the detected object(s). Analysis may include processing of sensor data from the autonomous vehicle to generate heat maps indicative of a location of the detected object(s) in the region during the time period. Probabilistic statistical analysis may be applied to the sensor data to determine a confidence level in the inferred semantic classification. The inferred semantic classification may be applied to the detected object(s) when the confidence level exceeds a predetermined threshold value (e.g., greater than 50%).

An information processing system according to an embodiment includes one or more hardware processors. The hardware processors acquire an n-dimensional vector. The hardware processors generate n coordinate arrays, where the n coordinate arrays is n pieces of n-dimensional arrays for which, with respect to each of elements of an m-th array (1≤m≤n), an element value having a same value as an index of an m-th dimensional coordinate of the elements is set. The hardware processors obtain n first probability distribution arrays including an output value of a probability density function as an element value corresponding to each of the n coordinate arrays, multiply n element values for each of elements corresponding to each of the n first probability distribution arrays, and obtain a second probability distribution array having a result of multiplication as an element value.