Abstract: A computer-implemented method includes receiving, by a computing device, a particular textual description of a scene. The method also includes applying a neural network for text-to-image generation to generate an output image rendition of the scene, the neural network having been trained to cause two image renditions associated with a same textual description to attract each other and two image renditions associated with different textual descriptions to repel each other based on mutual information between a plurality of corresponding pairs, wherein the plurality of corresponding pairs comprise an image-to-image pair and a text-to-image pair. The method further includes predicting the output image rendition of the scene.

Abstract: Disclosed is a system and method to automatically identify property-related risks through the use of computer vision, sensors, and/or building information models (BIMs). The ability to automatically identify a variety of hazards helps mitigate the associated risks, and thus reduces the number of accidents or fatalities that would otherwise occur. In some embodiments, a “risk map” can be generated by mapping the identified risks for a given property.

Abstract: Systems, methods, and apparatus for acoustic inspection of a surface are described. An example system may include an inspection robot structured to traverse an inspection surface in a direction of travel. The inspection robot may include a payload having a plurality of arms, connected to the inspection robot, to rotate around respective ones of a plurality of axes while the inspection robot traverses the inspection surface, where each of the plurality of axes is in the direction of travel. A plurality of sleds may be connected to the plurality of arms, and a plurality of inspection sensors connected to the plurality of sleds. The plurality of inspection sensors may be spaced apart from each other at adjustable positions to inspect the inspection surface at an adjustable resolution.

Abstract: A system including an inspection robot having a plurality of sensors, a further sensor, and a controller. The controller having circuitry to receive inspection data with a first resolution from the plurality of sensors, determine a characteristic on the inspection surface based on the inspection data, and provide an inspection operation adjustment in response to the characteristic, wherein the inspection operation adjustment includes a change from the first resolution to a second resolution. The change from the first resolution to the second resolution includes enabling the further sensor where the further sensor is at least one of: horizontally distributed with or vertically displaced from the plurality of sensors relative to a travel path of the plurality of sensors, and at least one of: offset in alignment from the travel path of the plurality of sensors, or operated out of phase with the plurality of sensors.

Abstract: A system includes an inspection robot comprising a plurality of sensor sleds; a plurality of ultra-sonic (UT) sensors; a couplant chamber mounted to each of the plurality of sleds, each couplant chamber comprising: a cone, the cone comprising a cone tip portion at an inspection surface end of the cone; a sensor mounting end opposite the cone tip portion; a couplant entry fluidly coupled to the cone at a position between the cone tip portion and the sensor mounting end; and wherein each of the UT sensors is mounted to the sensor mounting end of one of the couplant chambers.

Abstract: A system includes an inspection robot having an input sensor comprising a laser profiler and a plurality of wheels structured to engage a curved portion of an inspection surface, wherein the laser profiler is configured to provide laser profiler data of the inspection surface; a controller, comprising: a profiler data circuit structured to interpret the laser profiler data; determine a feature of interest is present at a location of the inspection surface in response to the laser profiler data; and wherein the feature of interest comprises a shape description of the inspection surface at the location of the feature of interest.

Abstract: Inspection robots and methods for inspection of curved surfaces with sensors at selected horizontal distances are described. An example of such an inspection robot includes a housing; a drive module with a wheel and a motor operatively linked to the housing, a plurality of sensor sleds, and a payload. The payload, which is coupled to the housing, may include a first and a second rail component, each with at least one connector, where the rail components are connectable at a first selected position of a plurality of discrete engagement positions. Each of the rail components may be structured to support at least one of the plurality of sleds where each of the plurality of sleds is coupled to the payload at a respective selected horizontal position such that the plurality of sleds are at selected horizontal distances from each other.

Abstract: System and methods for traversing an obstacle with an inspection robot are disclosed. An example system may include an inspection robot including an obstacle sensor to interrogate an inspection surface. The example may further include an obstacle sensory data circuit to interpret obstacle sensory data provided by the obstacle sensor, an obstacle processing circuit to determine refined obstacle data, and an obstacle notification circuit to generate and provide obstacle notification data to a user interface device. The example system may further include a user interface circuit to interpret a user request value from the user interface device, and to determine an obstacle response command value in response to the user request value; and an obstacle configuration circuit to provide the obstacle response command value to the inspection robot during the interrogating of the inspection surface.

Abstract: A system includes an inspection robot for performing an inspection on an inspection surface with an inspection robot, the apparatus comprising a position definition circuit structured to determine an inspection robot position on the inspection surface; a data positioning circuit structured to interpret inspection data, and to correlate the inspection data to the inspection robot position on the inspection surface; and wherein the data positioning circuit is further structured to determine position informed inspection data in response to the correlating of the inspection data with the inspection robot position, wherein the position informed inspection data comprises absolute position data.

Abstract: Systems and methods for an inspection robot having replaceable sensor sled portions are disclosed. An example system may include: an inspection robot including a plurality of payloads; a plurality of arms, each of the plurality of arms pivotally mounted to one of the plurality of payloads; and a plurality of sleds, each sled mounted to one of the plurality of arms. At least one of the plurality of sleds includes an upper portion coupled to a replaceable lower portion, where the replaceable lower portion includes a portion of a delay line for a sensor of the inspection robot.

Abstract: A system includes an inspection robot for performing an inspection on an inspection surface with an inspection robot, the apparatus comprising a position definition circuit structured to determine an inspection robot position on the inspection surface; a data positioning circuit structured to interpret inspection data, and to correlate the inspection data to the inspection robot position on the inspection surface; and wherein the data positioning circuit is further structured to determine position informed inspection data in response to the correlating of the inspection data with the inspection robot position, wherein the position informed inspection data comprises absolute position data.

Abstract: A system includes an inspection robot having a plurality of input sensors, the plurality of input sensors distributed horizontally relative to an inspection surface and configured to provide inspection data of the inspection surface at selected horizontal positions; a controller, comprising: a position definition circuit structured to determine an inspection robot position of the inspection robot on the inspection surface; a data positioning circuit structured to interpret the inspection data, and to correlate the inspection data to the inspection robot position on the inspection surface; and wherein the data positioning circuit is further structured to determine position informed inspection data in response to the correlating of the inspection data with the inspection robot position.

Abstract: A system includes an inspection robot comprising a lead inspection sensor providing lead inspection data, and a trailing inspection sensor; a controller, comprising: an inspection data circuit structured to interpret the lead inspection data; a sensor configuration circuit structured to determine a trailing sensor configuration change for the trailing inspection sensor in response to the lead inspection data; and a sensor operation circuit structured to adjust a trailing sensor configuration for the trailing inspection sensor in response to the trailing sensor configuration change.

Abstract: A system includes an inspection robot having a plurality of acoustic sensors coupleable to an inspection surface through a couplant chamber defining a delay line therebetween; the plurality of acoustic sensors configured to provide raw acoustic data; a controller, comprising: an acoustic data circuit structured to interpret the raw acoustic data; a thickness processing circuit structured to determine a primary mode value and a primary mode score value in response to the raw acoustic data; and wherein the thickness processing circuit is further structured to determine a thickness value in response to the primary mode value and the primary mode score value.

Abstract: A system includes an inspection robot having a plurality of input sensors, the plurality of input sensors distributed horizontally relative to an inspection surface and configured to provide inspection data of the inspection surface at selected horizontal positions; a controller, comprising: a position definition circuit structured to determine an inspection robot position of the inspection robot on the inspection surface; a data positioning circuit structured to interpret the inspection data, and to correlate the inspection data to the inspection robot position on the inspection surface; and wherein the data positioning circuit is further structured to determine position informed inspection data in response to the correlating of the inspection data with the inspection robot position.

Abstract: Disclosed is a system and method to automatically identify property-related risks through the use of computer vision, sensors, and/or building information models (BIMs). The ability to automatically identify a variety of hazards helps mitigate the associated risks, and thus reduces the number of accidents or fatalities that would otherwise occur. In some embodiments, a “risk map” can be generated by mapping the identified risks for a given property.

Abstract: Disclosed is a system and method to automatically identify property-related risks through the use of computer vision, sensors, and/or building information models (BIMs). The ability to automatically identify a variety of hazards helps mitigate the associated risks, and thus reduces the number of accidents or fatalities that would otherwise occur. In some embodiments, a “risk map” can be generated by mapping the identified risks for a given property.

Abstract: A system includes an apparatus for performing an inspection on an inspection surface with an inspection robot, the apparatus comprising: a controller configured to: interpret inspection data comprising sensed information from a location on an inspection surface; determine a feature of interest is present at the location of the inspection surface in response to the inspection data, and in response to determining the feature of interest is present at the location of the inspection surface, capture image information from the location on the inspection surface, and correlate the captured image information with the inspection data corresponding to the location of the inspection surface.

Abstract: A system includes an inspection robot, comprising a plurality of wheels that engage an inspection surface; a plurality of sensors positioned to interrogate the inspection surface; and wherein the plurality of wheels each comprise a first magnetic hub coupled to a second magnetic hub, and wherein the plurality of wheels further define a channel between the magnetic hubs.

Abstract: A system includes an inspection robot having a plurality of input sensors comprising a plurality of magnetic induction sensors and configured to provide inspection data of an inspection surface, wherein the inspection data comprises electromagnetic (EM) induction data, and wherein the plurality of input sensors are distributed horizontally relative to the inspection surface; wherein at least a portion of the inspection surface comprises a ferrous substrate having a non-ferrous coating thereupon; a controller, comprising: an EM data circuit structured to interpret the EM induction data, and to determine a substrate distance value in response to the EM induction data; and a thickness processing circuit structured to determine a thickness value in response to the EM induction data, the thickness value comprising a thickness of the non-ferrous coating.