A robotic device comprising a long-range sonar assembly configured to detect objects between 1000 mm to 9500 mm from the long range sonar assembly. The long-range sonar assembly comprises a flared bell housing and a transducer. The robotic device also comprises a transportation mechanism configured to move the robotic device in various directions in response to instructions from a processing device that is in communication with the ling-range sonar assembly and the transportation mechanism. The processing device may cause the long-range sonar assembly to, via the transducer transmit one or more pulses and receive the one or more pulses as echo pulses having reflected off an object in an environment. The processing device may then cause a transmitter to transmit instructions to the transportation mechanism to move the robotic device in the environment based on the received one or more pulses.

An apparatus such as a robot capable of performing goal oriented tasks may include one or more touch sensors to receive touch perception feedback on the location of objects and structures within an environment. A fusion engine may be configured to combine touch perception data with other types of sensor data such as data received from an image or distance sensor. The apparatus may combine distance sensor data with touch sensor data using inference models such as Bayesian inference. The touch sensor may be mounted onto an adjustable arm of a robot. The apparatus may use the data it has received from both a touch sensor and distance sensor to build a map of its environment and perform goal oriented tasks such as cleaning or moving objects.

Noise reduction in a robot system includes the use of a gesture library that pairs noise profiles with gestures that can be performed by the robot. A gesture to be performed by the robot is obtained, and the robot performs the gesture. The robot's performance of the gesture creates noise, and when a user speaks to the robot while the robot performs a gesture, incoming audio includes both user audio and robot noise. A noise profile associated with the gesture is retrieved from the gesture library and is applied to remove the robot noise from the incoming audio.

A method of collision localization on a robotic device includes obtaining audio signals from a plurality of acoustic sensors spaced apart along the robotic device; identifying, based on a collision being detected, a strongest audio signal; identifying a primary onset time for an acoustic sensor producing the strongest audio signal, the primary onset time being a time at which waves propagating from the collision reach the acoustic sensor producing the strongest audio signal; generating a virtual onset time set, by shifting a calibration manifold, based on the identified primary onset time, the calibration manifold representing relative onset times from evenly spaced marker locations on the robotic device to the plurality of acoustic sensors; determining scores for the marker locations based a standard deviation of elements in the virtual onset time set; and estimating a location of the collision based on a highest score of the determined scores.

A device and a method for measuring a predefined characteristic of an interaction capable of being present between an interaction device and a user comprises an input interface for receiving at least one temporal sequence of data representative of the interaction and a classification module, which is connected to the input interface and processes the temporal sequence of data to output a measure of the predefined characteristic. The classification module has been configured from temporal sequences of learning data marked with the presence or absence of the characteristic. Such a measuring device may be used for detecting the start and/or end of the interaction between the user and the interaction device.

A method controls a robot. One or more processors receive sensor readings from one or more sensors that are monitoring a human in real time, where the human is currently observing a robotic action by a robot, and where the robotic action is a physical movement performed by the robot. The processor(s) determine a cognitive state of the human while the human is observing the robotic action by the robot, and then adjust the robotic action being performed by the robot based on the cognitive state of the human.

Methods, systems and apparatuses are disclosed for non-destructively a substrate using ultrasound waves, and enhancing resolution of imaging created from ultrasound signals that are back reflected from a substrate surface second, or back surface by maintaining the incident angles of the ultrasonic beams at the substrate second surface such that the ultrasonic beams strike the substrate second surface at an angle that is substantially perpendicular to the complex geometric profile of the substrate second surface by supplying known spatial coordinates to the system to maintain the incident angles of the ultrasonic beams at a predetermined angle relative to the substrate second surface.

Ultrasonic ranging systems and methods that emit coded bursts and correlate transduced acoustical echoes of the bursts with a receive template characterizing a burst code to determine time-of-flight information use receive templates of time-variable length to improve short-range object detection. The template length is based on a time index measured from the start of the burst emission. The detection can account for a dead zone of transducer ringing following a burst. A time-variable gain that is also based on the time index can be applied to the correlated signal. The length and gain can be adjusted with reduced temporal frequency to reduce computation cost.

A mechanism is described for facilitating general purpose input/output data capture and neutral cache system for autonomous machines. A method of embodiments, as described herein, includes capturing, by an image capturing device, one or more images of one or more objects, where the one or more images represent input data associated with a neural network. The method may further include determining accuracy of first output results generated by a default neural caching system by comparing the first output results with second output results predicted by a custom neural caching system. The method may further include outputting, based on the accuracy, a final output results including at least one of the first output results or the second output results.

A robot that uses sensor inputs for attention activation and corresponding methods, systems, and computer programs encoded on computer storage media. The robot can be configured to compute a plurality of attention signals from sensor inputs and provide the plurality of attention signals as input to the attention level classifier to generate an attention level. If a user is paying attention to the robot based on the generated attention level, the robot selects a behavior to execute based on the current attention level, wherein a behavior comprises one or more coordinated actions to be performed by the robot.