465 results about "Robotics" patented technology

In various examples, a technique for generating simulation data includes generating, via one or more simulations, simulation data associated with operation of a first machine in an environment. The technique also includes determining a command to the first machine based at least on the simulation data and a goal associated with the first machine and updating the simulation data based at least on the command. The technique further includes storing the simulation data, the command, and the updated simulation data in one or more data records, and causing a second machine to perform one or more actions based at least on the one or more data records.

The invention discloses a robot system with tactile perception, and belongs to the technical field of tactile robots. The problem that an existing robot is poor in environment pressure sensing capacity is solved. Comprising a head module, an integrated vision-inertial navigation sensing module and a head controller, two side surfaces are provided with a pair of rotatable connecting ends, and the outer surface is provided with a tactile sensor; the head controller transmits data to the upper computer through the communication module; the driving body module comprises a plurality of modularized joints which are connected end to end, touch sensors are arranged on the two side faces of each modularized joint, and the adjacent modularized joints are orthogonally connected; the modularized joint is integrated with a joint controller and a joint steering engine; the tail module integrates a tail controller and a tail steering engine, and tactile sensors are arranged on the two side faces of the tail module. The tail controller carries out data transmission with the head controller through the adjacent previous joint controller in sequence, and the head controller receives a control instruction of the upper computer and carries out data transmission with all other controllers. According to the invention, the environmental pressure can be sensed in all directions.

An autonomous warehouse robotics system integrates a multi-sensor platform, adaptive payload handling, dynamic task reallocation, advanced navigation, energy management, and comprehensive safety features into one mobile robot chassis. The system utilizes LiDAR, stereo vision, ultrasonic sensors, mmWave radar, thermal cameras, and event cameras to perform complete environmental sensing and obstacle detection. Sensor fusion combines adaptive weighting, multi-modal data integration, and statistical filtering to create high-confidence maps for reactive path planning and collision avoidance. The robot's payload system features machine vision for item recognition, telescopic lifts, variable-width grippers, and real-time toolhead verification to handle a variety of goods. Fleet management is achieved through dynamic task reallocation that considers robot location, battery level, and operational delays, all coordinated by a cross-platform middleware architecture that ensures standardized communication, remote monitoring, and over-the-air updates among diverse robot brands. Energy management optimizes power usage via predictive routing, autonomous return-to-charge, and auction-based scheduling. Safety is maintained through proximity detection, behavior-based intervention, and human-robot cohabitation protocols while advanced localization is enhanced by fusing ultra-wideband positioning with visual landmark alignment, inertial sensing, and machine learning to deliver high accuracy in non-line-of-sight conditions. An onboard edge AI module further refines navigation and task prioritization through neural network inference, ensuring robust, adaptive operation in dynamic, unstructured warehouse environments.

A system for autonomous or semi-autonomous operation of a vehicle is disclosed. The system includes a machine automation portal (MAP) application configured to enable a computing device to (a) display a map of a work site and (b) provide a graphical user interface that enables a user to (i) define a boundary of an autonomous operating zone on the map and (ii) define a boundary of one or more exclusion zones. The system also includes a robotics processing unit configured to (a) receive the boundary of the autonomous operating zone and the boundary of each exclusion zone from the computing device, (b) generate a planned command path that the vehicle will travel to perform a task within the autonomous operating zone while avoiding each exclusion zone, and (c) control operation of the vehicle so that the vehicle travels the planned command path to perform the task.

A node-edge symbolic consent kernel (NESCK) provides a computing architecture in which every instruction is gated by a verifiable human-intent signal and an ethical-predicate chain prior to execution. The system integrates a biometric-sensing front-end (EEG / GSR / facial micro-affect), a symbolic arbitration engine that transforms bio-intent data into consent tokens, and a cryptographically bonded node-edge ledger that records execution lineage, revocation, and audit proofs. Each node represents an executable state bound to a human consent fingerprint, while each edge encodes the ethical transition rules authorizing propagation through the network. At runtime, the kernel evaluates symbolic predicates, verifies zero-knowledge proofs of consent, and allows or halts instruction dispatch. The framework operates across devices, edge nodes, and cloud layers, enabling real-time lawful AI behavior, revocable autonomy, and tamper-proof moral audit trails. Embodiments span neuroadaptive wearables, autonomous vehicles, robotics controllers, and sovereign AI systems requiring continuous consent and transparent accountability.

In various examples, a technique for neural motion control includes generating a first trajectory associated with a first end effector of an articulated object based at least on a first reference trajectory and one or more transformations between one or more reference poses for one or more objects and one or more new poses for the one or more objects. The technique also includes generating a second trajectory associated with a second end effector of the articulated object based at least on a second reference trajectory and the one or more transformations. The technique further includes determining that a task is successfully completed via execution of the first trajectory and the second trajectory and updating one or more parameters of a machine learning model based at least on the first trajectory and the second trajectory to produce a trained machine learning model.

The invention provides an indoor robot target detection method and system based on multi-modal tracking and a medium, and belongs to the field of robotics.The method comprises the steps that laser data processing is optimized through static furniture learning and filtering and dynamic region clustering, and the visual anti-illumination interference capacity is combined; a dynamic cost matrix matching strategy based on target motion characteristics is adopted to obtain a stable trajectory of a target person; kalman filtering trajectory prediction and target appearance feature matching are fused, and it is ensured that tracking can be rapidly recovered after the target is temporarily lost; a human-guided following mapping mode is adopted, a moving track of a tracked target is used as dynamic guidance, dynamic and static point clouds are separated, and a high-precision environment map with semantic tags is constructed; according to the invention, a following state monitoring and active service linkage mechanism is established, scene decision is carried out by identifying a target attitude and combining a semantic map, corresponding health monitoring and safety services are triggered, and the crossing from passive tracking to active intelligent services is realized.

In various examples, a technique for neural motion control includes determining a plurality of masks associated with one or more control modes for an articulated object. The technique also includes applying the plurality of masks to a plurality of goal state attributes for the articulated object to produce one or more masked goal state attributes. The technique further includes generating, via execution of a trained machine learning model, one or more actions based at least on the one or more masked goal state attributes and performing a task using the articulated object based at least on the one or more actions.

A comprehensive spatial AI platform for neural 3D reconstruction and built environment analysis integrates foundation models, deep learning methods, and spatial reasoning capabilities to provide expert knowledge of the physical world. The platform combines symbolic AI and machine learning to facilitate 3D semantics for insurance, real estate, construction, robotics, and other business applications. The system processes captured images, videos, or point clouds through neural networks with low compute requirements, incorporating device-agnostic advanced spatial intelligence that delivers geometric, semantic, and relational data. The platform includes proprietary training innovations, comprehensive measurement and semantic understanding, external sensor integration, human-in-the-loop quality assurance, and API integration for programmatic access. Advanced spatial reasoning capabilities enable property damage assessments, construction progress tracking, robotics navigation, and real-time applications including room dimension validation and automated repair estimates, supporting enterprise workflows across various industry segments while enabling productivity improvements and new value-added spatial AI use cases.

The invention provides a control structure of a rope-driven dexterous hand, the rope-driven dexterous hand and a robot, and relates to the field of robotics.The control structure comprises a master-slave driver, the master-slave driver comprises a master control chip and a plurality of slave control chips connected with the master control chip, each slave control chip is integrated with an FOC algorithm hardware acceleration module, and the FOC algorithm hardware acceleration module is connected with the master control chip. Wherein the number of the slave control chips is equal to that of finger driving motors of the rope-driven dexterous hand, the slave control chips and the finger driving motors are in one-to-one correspondence, and each slave control chip is connected to the corresponding finger driving motor through one three-phase full-bridge driver.

The embodiment of the invention provides power transmission line hardware fitting robot equipment, and relates to the field of power transmission line robots. The power transmission line hardware fitting robot equipment comprises a hoisting module and a robot module, the robot module is used for being fixed to a wire through an unmanned aerial vehicle, the robot module and the hoisting module are of an up-down split structure, the robot module and the hoisting module are detachably connected, and the robot module conducts autonomous online or offline through the hoisting module. An independent hoisting module is configured, and the hoisting module is conveyed and fixed to a wire through an unmanned aerial vehicle; the robot module achieves vertical lifting and accurate butt joint from the ground to the wire through the synergistic effect of the robot module and the hoisting module. The mode gets rid of dependence on personnel on the tower or an auxiliary sliding rail, the automation degree and safety of the on-line process are remarkably improved, and the method is particularly suitable for rapid deployment in a complex terrain or an electrified environment. And efficient, safe and autonomous online and offline of the robot module are realized.

In various examples, a technique for neural motion control includes applying a plurality of masks to a plurality of goal state attributes for an articulated object to produce one or more masked goal state attributes. The technique also includes generating, via execution of a machine learning model, one or more actions based at least on the one or more masked goal state attributes. The technique further includes updating one or more parameters of the machine learning model based at least on the one or more actions and one or more reference actions for the articulated object to produce a trained machine learning model.

The invention relates to the technical field of robots, in particular to a mobile robot remote operation system and method based on mixed reality, and the method comprises an operator end and a robot end, the operator end completes a complex remote operation task based on visual feedback provided by a man-machine cooperation interaction interface, and the robot end completes a remote operation task based on visual feedback provided by the man-machine cooperation interaction interface. The virtual reality controller integrates a motion state and an operation mode of the mobile robot, and the mixed reality server generates visual feedback according to visual information of a remote operation visual angle of the mobile robot; the robot end comprises a mobile robot, a collection device on the mobile robot and a robot server, and after the mobile robot receives a remote control instruction sent by the virtual reality controller to the robot end, posture and button information of the mobile robot are input into the robot server to drive the mobile robot to move. Therefore, the problems of low real-time performance and accuracy of remote control, poor operation efficiency and flexibility and the like caused by limitation of bandwidth and system time delay in related technologies are solved.

The invention relates to a mechanical arm disordered grabbing algorithm, and belongs to the technical field of robotics, and the method comprises the steps: firstly, obtaining a target material image and a point cloud sample, and training a multi-modal attention fusion perception model through sample enhancement processing and a difficult case mining mechanism after the target material image and the point cloud sample are labeled with categories, postures and grabbing areas; secondly, a mechanical arm obtains operation area data, the operation area data are input into a multi-modal attention fusion perception model to obtain initial attitude parameters, errors are corrected by combining with a visual data fusion calibration model, and the material stacking state is judged through a dynamic reference comparison method; thirdly, based on the correction parameters and the stacking state, an evaluation function is constructed to screen collision-free and stress-balanced grabbing points, and a collision-free path is planned by fusing kinematics constraints of the mechanical arm; and finally, the grabbing action is executed, the strength and the posture are adjusted in real time according to the contact force and the position deviation, the material stability is evaluated through vision after grabbing, and the stability and the operation efficiency of the mechanical arm in a complex disordered scene are improved.

Improved methods and a system for feeding containers and caps into a filling line and a capping line are provided. A 3D vision inspection system identifies a container located proximal to a top of a heap within a containers bin. A first set of robotic arms picks the identified container, places it onto a conveyor input in an upright orientation with the open end of the container facing upwards. The container is then transferred into a conveyor or accumulation table for transport to a filling station. Similarly, a third 3D camera identifies caps within a caps bin. A second set of robotic arms picks the identified caps, places them onto an alignment station, and orients them for placement into the capping line. The system incorporates Artificial Intelligence (AI)-based computer vision models for object identification and orientation, enhancing the reliability and efficiency of automated bottling and capping operations.

The invention relates to a high-efficiency three-dimensional point cloud classification method and device based on a pulse neural network, and the method obtains an event-driven point cloud expression form through sparse pulse voxel coding, and achieves the efficient extraction of local features through the combination of a pulse-driven sparse convolution structure. And a pulse self-attention module driven by logical operation is utilized to construct global correlation characteristics, and finally, category judgment is completed through a pulse time integration mechanism, so that the calculation amount and energy consumption are remarkably reduced while the classification precision is kept. According to the method, sparse activation and event driving characteristics of the pulse neural network are fully utilized, efficient processing of the three-dimensional point cloud is achieved, and the method is suitable for automatic driving, robots and embedded scenes sensitive to power consumption.

The invention provides a novel hand-eye calibration method based on finite geometric feature object reconstruction, and relates to the technical field of robotics and computer vision. The method comprises the following steps: firstly, constructing a hand-eye system, collecting point clouds of finite geometric feature objects, and defining coordinate systems and transformation matrixes of the hand-eye system; establishing a hand-eye calibration reverse iteration model based on each coordinate system of the hand-eye system, the conversion matrix and the hand-eye relation equation; correcting the initial poses of the point clouds, and adjusting all the point clouds to the lowest reachable poses capable of finishing point cloud precise registration; and finally, after the relative poses among the point clouds are obtained, solving the hand-eye calibration reverse iteration model to complete hand-eye calibration. The method can effectively replace a standard calibration object, the calibration cost is reduced, and the calibration flexibility is improved.

In various examples, a technique for performing a task includes converting one or more sensory inputs obtained using one or more sensors of a machine into a plurality of segments. The technique also includes, for each segment included in the plurality of segments, generating, via execution of a machine learning model, a caption for the segment, and storing, in a data store, a representation of the caption in association with the segment. The technique further includes performing, by the machine, one or more actions based at least on one or more queries of the data store.

In various examples, a technique for generating speed change decisions for a mobile robot includes identifying, using one or more maps of a physical environment, one or more obstacles associated with one or more portions of a path of the mobile robot in the physical environment. The technique also includes generating, based at least on the one or more obstacles, one or more speed constraints, each speed constraint specifying a speed limit for a respective portion of the path. The technique further includes generating one or more speed change decisions specifying actions to be performed by the mobile robot to cause a speed profile of the mobile robot to satisfy the one or more speed constraints.

The invention relates to the technical field of robots, and discloses a jewelry selling robot based on intelligent interaction, which comprises a machine body, and further comprises a frame body fixedly connected to the bottom of the machine body, a driving base fixedly connected to the bottom of the frame body, and a plurality of guide frames which are arranged at equal angles and are horizontally connected to the interior of the frame body in a sliding manner, the multiple pressed parts fixedly connected with the guide frame are arranged, secondary buffering stabilization is conducted through the stability maintaining part, and the stability maintaining part keeps stable continuously when the transmission part does not return to the initial state at the moment. Through the multi-layer structure, accurate coping with collisions with different strengths is achieved, buffer springs directly absorb energy during slight collision, frequent action of a transmission part is avoided, inclined grooves of guide grooves convert horizontal force into vertical force during large collision, all the buffer springs are compressed together, a magnetic ring and an adsorption ring are adsorbed stably, deviation of the robot is avoided, and the service life of the robot is prolonged. And the customer is prevented from being injured when being collided.

The invention discloses a robot control method, device, equipment, medium and product, and belongs to the crossing field of the robot technology and the artificial intelligence technology, and the method comprises the steps: constructing a teacher model and a student model based on a deep reinforcement learning algorithm, the first strategy network of the teacher model comprises a first basic action generation module and a first residual action generation module, and the second strategy network of the student model comprises a first basic action generation module and a second residual action generation module; a teacher model is trained based on the privilege information, so that the teacher model generates teacher actions, and the privilege information is used for being input into a first residual action generation module; and based on the teacher model, performing dynamic knowledge distillation on the student model, and adopting the trained student model to control the robot. According to the method, the problem that implicit state characterization between the student model and the teacher model is difficult to align during robot control and the problems of coupling of control signals and fuzzy learning targets can be solved.

The utility model relates to the technical field of robots, in particular to a wire harness fixing assembly and a robot. The wire harness fixing assembly is applied to the robot, the robot comprises a robot body, and the wire harness fixing assembly comprises a fixing piece and a fixing support. The fixing piece is provided with a wire fixing hole, the wire fixing hole penetrates through the two ends of the fixing piece along the designated axis, a limiting part is arranged on the periphery of the fixing piece, and the fixing piece has the circumferential direction surrounding the designated axis. The fixing support comprises a fixing part and a connecting part, the fixing part is used for being connected with a machine body, the connecting part is used for being detachably connected to the fixing part through the lacing part, the connecting part is provided with a positioning hole for the lacing part to penetrate through, the positioning hole and the limiting part are located in the same circumferential direction of the fixing part, and the positioning hole is used for being matched with the limiting part to limit the lacing part. The wire harness fixing assembly prolongs the service life of the wire harness.

A comprehensive spatial AI platform for neural 3D reconstruction and built environment analysis integrates foundation models, deep learning methods, and spatial reasoning capabilities to provide expert knowledge of the physical world. The platform combines symbolic AI and machine learning to facilitate 3D semantics for insurance, real estate, construction, robotics, and other business applications. The system processes captured images, videos, or point clouds through neural networks with low compute requirements, incorporating device-agnostic advanced spatial intelligence that delivers geometric, semantic, and relational data. The platform includes proprietary training innovations, comprehensive measurement and semantic understanding, external sensor integration, human-in-the-loop quality assurance, and API integration for programmatic access. Advanced spatial reasoning capabilities enable property damage assessments, construction progress tracking, robotics navigation, and real-time applications including room dimension validation and automated repair estimates, supporting enterprise workflows across various industry segments while enabling productivity improvements and new value-added spatial AI use cases.

A robotics corner module includes a steering drive unit, a traction drive unit, and an intermediate assembly. The steering drive unit includes an output shaft configured to be rotationally driven about an output shaft axis. The traction drive unit includes a wheel configured to be rotationally driven about a wheel hub axis. The intermediate assembly extends between the steering drive unit and the traction drive unit and includes a support structure and a cooling system. The support structure is operably coupled to the output shaft and the traction drive unit, such that rotation of the output shaft about the output shaft axis drives rotation of the intermediate assembly and the traction drive unit about the output shaft axis. The cooling system includes a radiator mounted to the support structure and a fan mounted to the support structure and configured to deliver air to the radiator.

There was a lack of understanding of cavity warning signs during normal times, and in the event of a disaster, advance road clearance plans could not be dynamically updated to reflect the actual situation, resulting in delays and personal delays. There was no integrated operation of authority adjustment across multiple administrators, allocation of equipment and personnel, preparation for communication outages, or reflecting knowledge from training and past performance. Solution: The information acquisition department collects data from satellites, optical fiber, vehicles, etc., and the analysis department performs integrated analysis. The road clearance department registers a preliminary plan, and the road clearance implementation plan is dynamically optimized based on the analysis results. The feasibility of achieving 24 / 48 / 72-hour targets is evaluated using reinforcement learning, and consideration is given to determining the transfer of authority, spatial optimization of bases and equipment, redundant communications, heavy equipment operation and robotics collaboration, training and performance learning, and conflict assessment through intellectual property investigations. Multi-hazard risks are quantified and weighted by priority, and a redundant network is formed in conjunction with the evaluation of bases such as roadside stations. Special equipment and sea and air transport are integrated, and abandoned vehicle removal and the automatic assembly of contracted companies are controlled, and emergency restoration plans are automatically generated along with alternative route selection.

Foundation models such as VLMs and LLMs are increasingly being used to address open-world tasks in robotics. While these models excel at task execution, they often face challenges in detecting and reasoning over failures, which are skills that are crucial for navigating dynamic and complex environments. The present disclosure provides a VLM that detects and reasons about failures in robotic manipulation using natural language, where the natural language feedback can then be used for improving downstream task performance through error correction.

The invention provides a semantic scene reconstruction and interaction method and system based on a visual language model, and relates to the technical field of artificial intelligence and robots. The method comprises the following steps: acquiring image data shot by a camera; estimating a world pose of the camera; inputting the image data to a Visual Language Model (VLM) to generate a structured description containing an object semantic tag and quantitative spatial information thereof relative to the camera; calculating the world pose of the object according to the camera pose and the quantitative space information; a node representative of the object is created or updated in a hierarchical scene graph. The invention further provides a corresponding interaction system. According to the method, a rich and visual three-dimensional semantic map facing natural language interaction is constructed by deeply fusing the semantic understanding ability of the VLM and the geometric mapping ability of the SLAM, and the intelligent level of robot environment perception and man-machine interaction is greatly improved.

The invention provides a self-adaptive dynamic current protection method and related equipment, and relates to the technical field of robots. Operation state information and task working condition information of a joint module are obtained, a current protection threshold value is dynamically determined according to the operation state information and the task working condition information, a future current peak value is determined in combination with prediction of a future operation state, then an optimal current protection threshold value is comprehensively determined, and finally graded protection of the joint module is achieved. The method can effectively solve the problem that in the prior art, a single fixed current protection threshold value cannot distinguish'benign overload 'and'fault overload', and the defect that when the robot executes a high-dynamic task, protection is triggered by mistake due to a transient current peak value, and consequently the performance of the robot is limited is overcome.