Aurora Operations, Inc.

Training and deploying a machine learning model in a low-latency runtime environment includes identifying a trained neural network including one or more parameters based on a specification of a pipeline element within a pipeline, the pipeline element corresponding to one or more operations executable by a central processing unit (CPU), the pipeline corresponding to a runtime environment having a low latency, the trained neural network generating a predicted output corresponding to a baseline output of the pipeline element in the runtime environment, sending an input associated with the pipeline to the trained neural network in the runtime environment, and generating the predicted output associated with the pipeline in the runtime environment based on the trained neural network, the trained neural network corresponding to one or more operations executable by the CPU.

A radio detection and ranging (RADAR) sensor system for vehicles, such as autonomous vehicles, includes a first RADAR sensor configured to provide first RADAR data descriptive of an environment of a vehicle having a first antenna configured to output a first RADAR beam having a first azimuthal component over a first angular range and a second RADAR sensor configured to provide second RADAR data descriptive of the environment of the vehicle, the second RADAR sensor having a second antenna configured to output a second RADAR beam having a second azimuthal component that is narrower than the first azimuthal component of the first RADAR beam, wherein the second RADAR sensor is configured to sweep the second RADAR beam over a second angular range closer to a rear of the vehicle than a front of the vehicle to obtain the second RADAR data.

An example computer-implemented method includes: obtaining environment data descriptive of one or more traffic signal devices of a traffic control node in an environment of an autonomous vehicle; generating a control node graph based on the environment data including vertices respective to representations of the traffic signal devices and edges indicative of relationships between the representations of the traffic signal devices; providing the control node graph as input to a control node graph processing model operable to reduce the control node graph to a distilled representation of the control node graph encoding information about a state of the traffic control node; based on receipt of the control node graph as input, generating an output based on the control node graph processing model; generating a motion plan based on the output from the control node graph processing model; and controlling the autonomous vehicle based on the motion plan.

Aspects of the present disclosure involve a vehicle computer system comprising a computer-readable storage medium storing a set of instructions, and a method for online light detection and ranging (Lidar) intensity normalization. Consistent with some embodiments, the method may include accumulating point data output by a channel of a Lidar unit during operation of an autonomous or semi-autonomous vehicle. The accumulated point data includes raw intensity values that correspond to a particular surface type. The method further includes calculating a median intensity value based on the raw intensity values and generating an intensity normalization multiplier for the channel based on the median intensity value. The intensity normalization multiplier, when applied to the median intensity value, results in a reflectivity value that corresponds to the particular surface type. The method further includes applying the intensity normalization multiplier to the point data output by the channel to produce normalized intensity values.

A light detection and ranging sensor system includes a laser source configured to output a source beam; a modulator configured to receive a modulation signal and modulate the source beam based on the modulation signal to produce a modulated beam; an amplifier configured to amplify the modulated beam; and a protection circuit configured to detect, by evaluating at least one of the modulation signal or a parameter of the modulated beam, a condition associated with the modulated beam; and control input of the modulated beam to the amplifier in response to detecting the condition.

Systems and methods are provided for forecasting the motion of actors within a surrounding environment of an autonomous platform. For example, a computing system of an autonomous platform can use machine-learned model(s) to generate actor-specific graphs with past motions of actors and the local map topology. The computing system can project the actor-specific graphs of all actors to a global graph. The global graph can allow the computing system to determine which actors may interact with one another by propagating information over the global graph. The computing system can distribute the interactions determined using the global graph to the individual actor-specific graphs. The computing system can then predict a motion trajectory for an actor based on the associated actor-specific graph, which captures the actor-to-actor interactions and actor-to-map relations.

Systems and methods for emergency vehicle detection are provided. An example method includes obtaining sensor data including image frames indicative of an actor in an environment of an autonomous vehicle. For a respective image frame, the example, method includes determining, using a machine-learned model, that the actor is an emergency vehicle and generating output data indicating that the emergency vehicle is active or inactive. The example method includes storing (e.g., in a buffer) attribute data for the respective image frame. The attribute data includes the output data and a time associated with the respective image frame. Once the buffer reaches a particular threshold, the example method includes determining (e.g., using a second model) that the emergency vehicle is an active emergency vehicle based on the attribute data. The example method includes performing an action for the autonomous vehicle based on the active emergency vehicle being within the autonomous vehicle's environment.

An apparatus include a motor, a first scanner, and a second scanner. The first scanner is coupled to the motor, and the motor is configured to rotate the first scanner at a first angular velocity about a rotation axis to deflect a first beam incident in a third plane on the first scanner into a first plane different from the third plane. The second scanner is coupled to the motor, and the motor is configured to rotate the second scanner at a second angular velocity different from the first angular velocity about the rotation axis to deflect a second beam incident in the third plane on the second scanner into a second plane different from the third plane.

A sensor cleaning system for an autonomous vehicle includes at least one sensor configured to generate a sensor signal. The at least one sensor includes at least one sensor lens. At least one reservoir for a fluid and at least one conduit fluidly coupled to the at least one reservoir and to at least one nozzle are configured to direct the fluid towards the sensor lens. At least one processor is communicatively coupled to the at least one sensor. A memory is operably coupled with the at least one processor, where the memory stores instructions that cause the at least one processor to perform operations including: receiving the sensor signal from the sensor; comparing a detected quality of the sensor signal relative to a nominal quality of the sensor signal; and, dispensing the fluid from the at least one reservoir through the nozzle towards the sensor lens.