patent.fyi
Applications  /  US 2026/0355949
United States Patent Application Publication

US 2026/0355949 A1

Published
Nov 12, 2026
Filed
Jul 8, 2026
Earliest date on the face
Feb 6, 2024
Classification
B25J 9/163
First published here
Nov 19, 2026

Generated automatically. Not reviewed by a person. Not legal advice or an opinion.

Section 1 · machine-generated

What if?

Questions about how individual parts of claim 1 might be changed. A question may describe something different from what the claim language recites, or a version of it; no view is offered on which.

1.1Timing from actuator limits

What ifWhat if a deterministic planner computes waypoint arrival times from actuator limits?
Claim language“trained based on first sample trajectory data”

The planner maps the path into joint space and computes a speed profile from joint speed and acceleration limits using forward and backward passes. It integrates travel time along the profile and records the cumulative time at each waypoint.

1.1Training from torque margins

What ifWhat if the training records pair operating conditions with measured torque margins?
Claim language“respective sample coordinates and respective sample timestamps”

Each record contains a payload mass, a path curvature, a commanded speed, and the difference between the measured peak joint torque and the joint torque limit. A regression model estimates the torque margin for candidate speeds, and a timing planner uses the resulting speed limits to compute waypoint arrival times.

1.2Path from an analytic curve

What ifWhat if the motion program supplies a helix equation and a parameter interval?
Claim language“respective first coordinates of at least two first path points”

The program stores a helix radius, a pitch, a reference frame, and a start and end angle. A planner computes progress along the helix, and the servo interpolator evaluates the curve at the current angle to generate a tool position each cycle.

1.3Motion driven by station events

What ifWhat if the motion data supplies station event codes that release the robot to move between path points?
Claim language“respective first timestamps corresponding to the at least two first path points”

Each station record pairs a target position with an input condition, such as a fixture-ready contact. The controller advances to the next position when the input condition is true and the robot is within the position tolerance of the current station.

1.3Arrival times entered by an operator

What ifWhat if an operator enters the arrival times directly into a waypoint table?
Claim language“determining, through the path parameterization model, first trajectory data”

The controller pairs each entered time with the coordinates in the corresponding row and checks that the times increase in path order. It generates polynomial segments between the positions and checks each segment against the joint speed and acceleration limits before execution.

Section 2 · machine-generated

Variations

Further arrangements of the same subject matter, described here as of the publication date at the foot of this page.

Other field of use

Timing passes for robotic surface inspection

A robot carries an ultrasonic probe on a spring-loaded mount and follows waypoints on a curved panel. A timing model trained on waypoint coordinates and probe arrival times from recorded inspection passes assigns a timestamp to each waypoint on the panel. The motion controller interpolates between the timed waypoints, limits tangential speed, and adjusts the probe along the surface normal to maintain a specified contact force. The controller triggers a measurement when the probe reaches each waypoint within its position tolerance and records the actual arrival time.

Component substitution

Nearest neighbor timing from recorded paths

A k-nearest-neighbor regressor serves as the path parameterization model and stores training records derived from timed paths. Each record contains the segment length, the turn angles at both ends, and the elapsed time between its endpoint timestamps. For each requested segment, the regressor retrieves five records by Euclidean distance in the feature space and uses their median elapsed time as the segment duration. The controller accumulates those durations into waypoint timestamps and stretches intervals where interpolation exceeds joint velocity or acceleration limits.

Parameter or scale variation

Dense timing for short positioning strokes

A linear manipulator follows a 20 mm stroke defined by 21 waypoints at 1 mm spacing. A timing model trained on measured coordinates and arrival times from repeated strokes assigns a timestamp to each waypoint. A controller interpolates the timed positions with a cubic spline and sends setpoints to the servo at 1 kHz. With both endpoint velocities set to zero, the controller expands the time intervals if the spline exceeds 100 mm/s or 500 mm/s².

Combination with existing systems

Waypoint timing synchronized with a conveyor

A belt encoder feeds a conveyor tracking controller that maintains a moving coordinate frame for each part. A timing model trained on coordinates and arrival times from recorded picks assigns timestamps to approach, grasp, and lift waypoints expressed in that moving frame. At each servo cycle, the tracking controller transforms the interpolated waypoint position into robot base coordinates using the current encoder count. The cell controller starts the pick when the grasp point falls within the robot's reachable window at the predicted grasp time and checks joint speed limits against the combined belt and arm motion.

Automated or data-driven variant

Waypoint timing adjusted by measured load

A force sensor at the wrist supplies a stationary measurement after the gripper lifts the workpiece, and the controller subtracts the gripper tare to estimate payload mass. A timing model trains on waypoint coordinates, arrival timestamps, and measured payload masses from recorded transfers. The model uses the current mass and the requested waypoint coordinates to assign arrival timestamps. A robot dynamics check evaluates joint torque along the interpolated motion and stretches any interval that exceeds the drive's torque ceiling.

Waypoint timing corrected by encoder feedback

A timing model trained on coordinates and arrival timestamps from recorded runs provides the initial schedule for a fixed waypoint path. Joint encoders feed forward kinematics, and the controller projects the measured tool position onto the current path segment to estimate progress. When measured progress falls more than 20 ms behind the schedule, the controller regenerates timestamps for the remaining waypoints from the measured position and velocity. The controller constrains the regenerated schedule to the configured joint velocity and acceleration limits and keeps the remaining waypoint coordinates fixed.

Specific version of a part

Training from recorded robot runs

A recorder samples joint encoders at 250 Hz and uses forward kinematics to compute tool positions. For each run, it records the position and elapsed time at each waypoint crossing, then uses the ordered records to train a recurrent regression model.

Waypoints from a pendant

An operator jogs the manipulator to each path point and presses a pendant key to record the tool position in the workpiece frame. The controller stores the ordered coordinate triples and joins successive points with straight segments.

Timestamps from positive segment durations

A sequence model receives the ordered waypoint coordinates and outputs a duration for each segment using an output function with a positive lower bound. The controller assigns time zero to the initial waypoint and cumulatively sums the durations to obtain the remaining arrival times.

Timed motion with joint limits

A trajectory interpolator generates joint setpoints at 1 kHz from the path and waypoint timestamps. Before motion begins, the controller checks interpolated joint positions and peak velocities against configured limits and enables execution only when both checks pass.

Section 3 · quoted from USPTO records

Other work

Passages reproduced word for word from two earlier-published U.S. patent documents, placed beside the language of claim 1. No characterization is offered. Each passage is marked A or B for the document it comes from; ¶ numbers are the paragraph numbers in the USPTO’s published full text of that document.

Document AUS 2017/0221368 A1
Published
Aug 3, 2017
Filed
Apr 21, 2017
Class
G08G 5/0039
Document BUS 9,975,244 B1
Published
May 22, 2018
Filed
Aug 2, 2016
Class
B25J 9/1664
Document CUS 10,828,775 B2
Published
Nov 10, 2020
Filed
Aug 31, 2018
Class
B25J 9/163
Claim 1
Document A
Document B
1.1acquiring a path parameterization model, the path parameterization model being trained based on first sample trajectory data, the first sample trajectory data including respective sample coordinates and respective sample timestamps of at least two sample path points in a sample path, and each of the sample timestamps indicating a sample path point moment at which a training object reaches a respective one of the at least two sample path points in a process of passing through the sample path;
That is, acquiring route data of the aerial vehicle may comprise acquiring the coordinates, heights, and time values of various waypoints of the aerial vehicle in the course of flight as the route data of the aerial vehicle.
A · US 2017/0221368 A1 · ¶ 44
1.2acquiring a first path and respective first coordinates of at least two first path points in the first path;
In some embodiments, acquiring the route data comprises acquiring a geometric figure drawn on a planar map, calculating slopes corresponding to position points on the geometric figure, determining two or more of the position points on the geometric figure as the waypoints on the route, and obtaining the route data according to coordinates of the two or more of the position points.
A · US 2017/0221368 A1 · ¶ 10
In some implementations, a path planner or other component may provide a target motion state that indicates end effector 106 should transition from its current motion state to a motion state where the reference point 108 is positioned at waypoint 103, such as a target motion state where the reference point 108 is positioned at waypoint 103 with the end effector 106 at zero velocity.
B · US 9,975,244 B1 · ¶ 41
1.3determining, through the path parameterization model, first trajectory data based on the first coordinates, the first trajectory data including respective first timestamps corresponding to the at least two first path points, and the first trajectory data being configured to control a robot to move on the first path such that the robot reaches the at least two first path points at respective first path point moments indicated by the first timestamps; and
In some embodiments, the method in accordance with the present disclosure may also comprise, when a flight control operation is detected, controlling the aerial vehicle to fly automatically according to time values, waypoint coordinates, and waypoint heights included in the updated route data.
A · US 2017/0221368 A1 · ¶ 70
1.4controlling the robot to move on the first path based on the first trajectory data and a constraint condition associated with the first trajectory data.
Generating the updated trajectory includes applying the new acceleration constraints to the real-time generator in lieu of the initial acceleration constraints. The method further includes providing the updated trajectories for operating the actuators of the robot in lieu of the trajectories.
B · US 9,975,244 B1 · ¶ 11

An empty cell means only that no passage was reproduced there.

Statements in these documents
For the automatic control, inputted data could not be fed back to the user intuitively.
A · US 2017/0221368 A1 · ¶ 5
The embodiments of the present disclosure provide a method and device for setting a flight route capable of configuring automatically a route display interface for a user to configure a certain route.
A · US 2017/0221368 A1 · ¶ 6
Additionally, there may be some configurations of a robot where the defined kinematic constraints may not be valid, and the robot may not be able to achieve a generated trajectory due to the dynamics of the robot in that configuration making it infeasible to track the planned motion.
B · US 9,975,244 B1 · ¶ 4
The present disclosure is generally directed to methods and apparatus for generating updated robot actuator trajectories in response to violation of torque constraints and/or other constraints in previously generated robot actuator trajectories.
B · US 9,975,244 B1 · ¶ 5
Cite aspatent.fyi, US 2026/0355949 A1, claim 1, first published 2026-11-19T17:33:15-05:00, https://patent.fyi/US20260355949/, SHA-256 5de734845c990cd7d6f4e22c6a5fb1d9dffb68f625220427a928e8947f270453
First published
Permanent addresshttps://patent.fyi/US20260355949/
Content fingerprint (SHA-256)5de734845c990cd7d6f4e22c6a5fb1d9dffb68f625220427a928e8947f270453
Methodv2026-11-18.2 · automated, not reviewed by a person
Correctionscorrections@patent.fyi