Alternating local overlap scans with wider-area scans updates 3D sensor maps faster while preserving precise coverage in robot work areas.
Selective robot emotional expression uses recognition difficulty, decision difficulty, and user attention to boost attachment while limiting power use.
Spray-painted cut lines and 3D curve reconstruction let a mobile robotic cutter follow irregular salvage metal paths with less worker exposure.
A 3D fiducial on the robot end-effector enables faster extrinsic camera calibration without removing attachments or changing the workspace.
Ranks 3D carton pose hypotheses by surprisal to unload mixed carton piles without feature tuning, retraining, or pre-known arrangements.
Sensors capture weight, rigidity, and center of gravity so robots and conveyors can adjust force and speed to cut picking errors.
Sensor-guided robots identify item attributes, plan grasps, and coordinate paths to separate mixed parcels with higher throughput and less labor.
A spherical adaptor lets ballbars address the tool center point directly, speeding robot TCP calibration without laser tracking.
A snake-arm robot threads through engine passages, places a servicing tool precisely, then detaches so the tool stays in place for reuse.
Geometric alignment of end effector footprints with segmented pick surfaces improves robotic pick pose accuracy in chaotic scenes with lower compute.
A 3D sensor and virtual robot model remove self-point data, improving detection of approaching objects without restricting robot motion.
Adjustable safety zones use cart or topper dimensions to cover overhangs and trigger controlled stops when nearby objects create collision risk.
A pose- and load-adaptive notch filter in each motor drive cuts multi-axis oscillations, reducing settling time while preserving positioning accuracy.
Programmable motion and barcode-based perception route varied objects to assigned stations, boosting throughput while reducing bins, space, and labor.
Map-based tags and 3D video overlays help telepresence robots navigate complex spaces with more precise remote control and context-aware actions.
Balances safety, hard, and soft control constraints to keep human-robot distance safe while preserving resumable trajectory tracking.
When workspace conditions change unexpectedly, the controller checks task feasibility, sets an abstract state, and regenerates executable subtasks.
Force and torque sensing plus AR guidance turn the manipulator arm into an intuitive control interface, reducing training and manual reliance.
An accelerating unit raises rotation speed before braking, cutting required torque so the brake can be smaller while preserving braking ability.
Permanent magnets in the trocar and arm-mounted sensors enable ML-based pose estimation for accurate robotic docking through sterile barriers.
Transfer learning with Gaussian processes reuses source-model knowledge to control physical or chemical processes with less data and lower computation.
An emulator mimics aircraft assembly jig interfaces to validate module signals early, reducing collision risk and setup delays.
Flexible suspension and cavity-protrusion fastening secure a robot arm cap without screws, avoiding dust buildup in clean room joints.
Workspace indicators let imaging and robotic frames be registered even when the end effector is obscured, avoiding extra hardware or reconfiguration.
Predictive strike planning and acoustic feedback help straighten hardened metal workpieces with fewer impacts and less material damage.
Multi-point deviation measurements and transformation-matrix offsets let the robot correct part pose drift and keep assembly within tolerance.
Deep learning object detection cuts setup and maintenance effort while reliably inspecting moving manufactured parts in real time.
Planned robot motions are treated as dynamic obstacles to generate collision-free paths in shared workspaces with high throughput.
When execution fails after object or environment changes, decentralized robots detect deviations and re-plan locally to finish tasks reliably.
A removable frame section lets the actuator rotate beyond 360° during calibration, improving rotation sensor accuracy without interference.
A projection-to-receiver coupling passes force through a continuous sterile barrier, keeping non-sterile drive parts isolated from the manipulator.
Automatic tangent-based curve selection connects 3D CAD spline segments into continuous robot path teaching data with less manual input.
Mutual-information-based control selection cuts GP state space model training time and computation for robotic device control.
Segmented blocks, linear actuators, and joint connectors deliver multi-DOF body assistance without adding rigid bulk or restricting bending and twisting.
Checkpoint logs with sensor IDs and timestamps trace robotic decision flows, exposing latency sources for diagnosis, training, and safe-mode response.
Sensor feedback switches the robot between high-speed and low-speed modes to reduce human interference and collision risk.
Joint dynamics and odometry help distinguish swing-leg touchdown from trips, enabling faster contact detection and better robot stability.
Cloud ASR and a dedicated operator channel filter multi-user speech so telepresence robots execute the right commands and save battery.
Automatic self-calibration updates industrial robot force sensor correction coefficients during operation to compensate drift and maintain precise force control.
A movable upper extremity support shifts outside the work area when assist is off, reducing interference while preserving gravity-counteracting arm support.
A signal conversion module maps game controller inputs to robot commands, enabling precise low-cost remote operation over a network.
Geometric modeling fused with vision data improves stack estimation for heterogeneous pallet loads, enabling stable, dense placement with lower-precision robots.
Goal and constraint regions keep high-level route context available locally, so robots can adapt to dynamic obstacles without leaving safe bounds.
A flange-mounted protection element carries sensors beyond the effector to detect concealed collision risks around robot workpiece handling.
Liquefied gas boosts pneumatic actuator output while cooling the generator, helping wearable robots deliver force without bulky hydraulic hardware.
Vision-guided vacuum gripping reorients mixed items for label visibility while reducing mishandling and damage at high handling speeds.
Image-based pallet detection lets a robot line divert recyclable yarn spindle pallets automatically, cutting manual sorting time and labor.
Graphical blocks gain editable text fields so robot programs can stay easy to build while still expressing advanced commands.
Sparse optical scans are corrected with sanding-head force feedback to align the workpiece model and improve finishing accuracy.
Two safe release signals and actuator selection prevent unintended MwoDP activation in industrial robot brake release control.