Decoupled horizontal and vertical leg motion helps robots avoid obstacles, meet touchdown timing, and maintain balance in constrained spaces.
Force and torque sensing lets an operator hand-guide a robot spray head, cutting programming effort while preserving precise repeatable surface treatment.
Hybrid grasp training combines simulation with sensor-validated real picks to improve object picking accuracy, speed, and runtime adaptation.
Wi-Fi access point IDs and signal strengths let robots correct position quickly in structure-poor spaces without costly LiDAR.
Links motion-capture arm paths with timed work inputs so robots can teach gripping and separation tasks with less correction.
Axial carrier forcing lets a robotic planetary gear train switch contact modes to cut backlash, friction losses, and tolerance sensitivity.
Fluid pressure replaces spring return in a collet handling unit, preserving coaxial alignment while reducing size and improving repeatable gripping.
AI learns airport user movement to detect wandering passengers and dispatch limited robots directly where navigation help is needed.
A retractable linear member lets multi-joint limb portions fold quickly into a compact shape for damage protection, storage, and transport.
Predictive 3D occupancy envelopes restrict robot motion from real-time human proximity, improving collaborative safety without fixed zones.
Onboard sensing updates stored 3D spatial information in real time, enabling safe robot inspection without pre-measured maps.
Using image capture and indication lamp reflections, the robot identifies mirrors and updates their map locations to avoid collisions.
A robot-mounted camera creates a virtual image sensor to calibrate the workpiece frame in one step, avoiding blockage, contact damage, and human error.
Rotating guide elements replace sliding cable friction, enabling efficient torque transmission between pulleys across changing axis orientations.
Automatic reference-position detection calibrates a pivoting operation unit without obstacle interference or manual input errors.
By linking robot motion with coating-state data in real time, this case improves coating precision, monitoring, and cross-device control.
Multi-modal perception combines visual and depth data to identify mixed objects, choose secure grasps, and route them with less manual sorting.
By linking robot motion data with coating states, this case improves defect detection and stabilizes coating quality with faster analysis.
Semantic obstacle classes let robots adjust speed, spacing, and path policies to avoid collisions and respect personal space.
A threaded tendon fastener replaces fixed knots with reversible tension and length adjustment for more reliable tendon-driven systems.
Movable stacking-bay walls and multi-axis compression form uniform pipe insulation bundles without vacuum damage, improving loading and shipment.
Intersecting bevel gears and three motors create a compact common-center rotation mechanism for precise multi-axis robot head movement.
Distributed sensors classify occupied and unknown regions and forecast motion paths to shrink off-limits space in human-robot workcells.
Small aligned support contacts stabilize the workpiece during robotic fastening, enabling precise multi-side attachment without protective cages.
Predicting time, speed, and energy for alternative robot paths helps select collision-free multi-pick sequences with higher pick rates.
Vision-guided grasp selection and motion planning automate sorting of mixed objects in cluttered, changing environments.
Coordinate conversion programs let one controller translate diverse interface commands into joint-axis targets for flexible multi-axis robot programming.
A ring network separates synchronous and asynchronous robotic arm data to cut latency and keep surgical control reliable.
Independent linear link motion is converted into precise 2-DOF rotation using parallel leaf springs, avoiding wire maintenance and easing sterilization.
Electromagnetic scan coordinates and vision-detected bin markings guide robotic baggage opening and OOI swab inspection with less delay.
UV-reflective invisible indicia let robots identify objects, navigate, and follow interaction instructions without disrupting human-facing aesthetics.
A right-angle bevel gear path and detachable housing cut motor projection, preserve internal space, and simplify robot joint assembly.
A robot grips, scans, then temporarily releases and regrips objects to shift position, improving storage efficiency and task accuracy.
When a robot hits a task bottleneck, subtasks are sent for targeted human input so work can resume and assistance data can improve autonomy.
A 3-DOF arm positions a container with an inclined polyurethane surface to catch bent pipes without damage or complex robot programming.
Stored form patterns let multiple support robots reposition parts for different 3D product shapes without dedicated jigs or long setup.
Sensor-generated 3D point clouds let a welding robot track seam shifts and update motion and weld settings in real time.
A robot-controlled spatula scrapes undried joining material from honeycomb segment joints to cut manual labor and prevent sinks and cracks.
A telescoping vacuum conduit keeps suction sealed on irregular objects while reducing tubing load, kinking, and robot motion interference.
A mobile robot follows the user to capture gait data from multiple positions, overcoming the limited coverage of fixed analysis setups.
Geometry-aware 3D encodings from 2D images improve robot grasp outcome prediction and reduce reliance on extensive real-world training.
Repeating image features are clustered to locate accurate robot gripping positions without time-consuming teaching models, even for piled objects.
Modified Lagrange multipliers add virtual spring-damper forces to stabilize coupled admittance controllers and simulate capped break-out behavior.
A defined bearing-to-flexible-gear diameter gap maintains lubrication and spreads load in wave gearing, improving robot gearbox life.
An elastic body redirects a surgical actuation cable path as slack develops, preserving tension and force transmission without complex adjustment.
A unified backend authenticates multiple users and robots, enabling remote surveillance control with local data storage and lower latency.
Neural networks turn tactile gripper data into sliding-velocity control, enabling gentle object placement in cluttered spaces.
Deep learning, simulation, and reinforcement learning help heterogeneous robots share logistics tasks with better allocation and resource use.
Polyhedral motion cone approximations enable stable 3D frictional pushes in a gripper while cutting computation for object repositioning.
Load torque detection and sensor-based arm alignment automate parallel robot calibration, improving accuracy while avoiding manual setup.