US 12287650 B1Patent grantUnited States

Adaptive Speed Control for Line-Following Robots

Official patent title

Adaptive speed control for line-following robots and method thereof

Arabic title: تحكم تكيفي في سرعة الروبوتات المتتبعة للخط وطريقته

Invention

Invention

Problem

IR-only line-following robots see little of the path ahead, so noise, sharp turns, intersections, lighting changes, and high speed can cause delayed correction or path deviation.

Why it matters

Anticipating line geometry before reaching a turn can support smoother and more controlled motion in assembly, warehouse, agricultural, and educational robot systems.

Approach

The robot combines a forward camera and IR sensor array, predicts line type with a neural network, computes a base speed, and uses PID-generated PWM signals to adjust left and right motors.

Who may benefit

Potential beneficiaries include factory-automation teams, mobile-robot manufacturers, warehouse operators, agricultural robotics groups, educators, and embedded-control researchers.

Potential value

The design joins forward-looking vision, close-range IR positioning, neural line classification, PID speed correction, and a compact geared four-wheel drivetrain in one controller.

Background

Background

Traditional line followers often use downward-facing IR sensors that detect only the line immediately beneath the robot. This narrow view limits advance warning of curves and intersections, and sensor noise can destabilize steering at higher speed. Camera systems provide a forward view but may add processing load, dependence on lighting, or bulky computing hardware. The patent combines both sensing modes: images classify the upcoming line type and determine a base speed, while the IR array estimates current position and a PID controller corrects motor speed through PWM.

Technology overview

Technology overview

A front camera supplies line images for an adaptively processed training set and a neural network that predicts line type. An arc-shaped array of IR reflection switches estimates base position. The controller calculates base speed from line type and may also consider surface friction and inclination. A PID algorithm converts speed and positional error into separate PWM signals for left and right motors. Each motor drives two wheels through coplanar herringbone gears, and the system can detect a finish line and stop. Operator input and a status display are also claimed.

Potential applications

Potential applications

  1. Potential use in assembly-line material-transport robots.
  2. Potential use in warehouse line-guided mobile platforms.
  3. Potential use in agricultural or event-navigation robots on marked paths.
  4. Potential use in robotics education and adaptive-control research.

Evidence-supported advantages

Evidence-supported advantages

  1. Uses forward vision to anticipate line type before the IR array reaches it.
  2. Combines neural classification with a conventional PID control loop.
  3. Provides independent PWM adjustment of left and right motor speeds.
  4. Includes finish-line stopping, operator input, and displayed status information.

Development stage

Development stage

Patent publication describing a robot architecture and control workflow; independent prototype performance, operational deployment, safety validation, and commercialization were not established.

Commercial opportunity

Commercial opportunity

The system may support licensing or prototyping with mobile-robot and factory-automation developers. Validation should benchmark speed, turn success, line-loss recovery, lighting variation, surface friction, slope, payload, camera latency, model generalization, controller stability, braking distance, obstacle safety, mechanical wear, battery life, and compliance with industrial mobile-robot requirements.

Patent classifications

Patent classifications

WIPO IPC

  • G05D1/65Systems for controlling or regulating non-electric variables
  • G05D1/646Systems for controlling or regulating non-electric variables
  • G05D101/15Systems for controlling or regulating non-electric variables

CPC

  • G05D1/646Systems for controlling or regulating non-electric variables
  • G05D1/244Systems for controlling or regulating non-electric variables
  • G06V20/56Image or video recognition or understanding
  • G05D1/65Systems for controlling or regulating non-electric variables
  • G06V10/82Image or video recognition or understanding
  • G05D2111/14Systems for controlling or regulating non-electric variables
  • G05D2107/70Systems for controlling or regulating non-electric variables
  • G05D2109/10Systems for controlling or regulating non-electric variables
  • G05D2101/15Systems for controlling or regulating non-electric variables

Inventors

Inventors

  • First inventorAbdul Khader Jilani Saudagar
  • InventorSafkat Shahrier Swapnil
  • InventorSandip Kumer Sarker
  • InventorArpon Bose Dibya
  • InventorMd Tanvir Islam
  • InventorMd. Abu-Talha Roni
  • InventorHitoun A. Alsagri
  • InventorKhan Muhammad

Keywords

Keywords

  • line-following robot
  • adaptive speed
  • computer vision
  • IR sensor array
  • neural network
  • PID control
  • PWM
  • assembly line

Patent document and drawings

Patent document and drawings

The patent publication is mapped to this record. Patent drawings remain within that publication; no separately cleared public media package has been supplied.