Path Planner Robot for Indoor Positioning using RF Ranging GPS

Path Planner Robot for Indoor Positioning using RF Ranging GPS

How It Works

• Three UWB anchor modules (e.g. DecaWave DWM1000) are fixed at known coordinates around the operating space, forming a reference grid.
• The robot's onboard UWB tag exchanges time-of-flight ranging messages with each of the three anchors; the resulting distances are trilaterated into the robot's real-time (x, y) position, accurate to roughly 10cm.
• The user enters a target coordinate through an Android app, sent to the robot over Bluetooth/Wi-Fi.
• An onboard path-planning routine computes a route from the robot's current UWB-derived position to the target.
• A magnetometer (digital compass) cross-checks and corrects heading drift during travel, since wheel odometry alone accumulates error over a route.
• Mecanum wheels let the robot strafe and rotate precisely to align exactly with the target coordinate without needing a wide turning arc.

Components

Arduino Mega or ESP32 (robot controller)
3x UWB DWM1000 anchor modules (fixed, known coordinates) + 1 mobile UWB tag (on robot)
Magnetometer (HMC5883L / QMC5883L)
4x Mecanum wheels + DC gear motors
Dual motor driver modules (L298N / BTS7960) for 4-wheel independent control
Bluetooth (HC-05) or built-in Wi-Fi (ESP32) for Android app communication
Android app for target-coordinate input
Battery pack + chassis

Applications

• Warehouse and indoor delivery robots
• Hospital equipment/medication transport bots
• Indoor mobile robot and drone navigation research
• GPS-denied environments — basements, tunnels, multi-storey buildings

Advantages

• ~10cm UWB accuracy vs GPS's 20-50m — and works fully indoors, where GPS doesn't reach at all
• App-based target-coordinate input replaces manual driving with a simple destination command
• Magnetometer heading correction keeps the robot on course over longer routes than wheel odometry alone
• Mecanum wheels give precise final positioning without wide turning maneuvers

Sample Code — UWB Trilateration with Magnetometer Heading Correction

// Simplified UWB trilateration + heading correction outline - Arduino
struct Point { float x, y; };
Point anchor1 = {0, 0}, anchor2 = {500, 0}, anchor3 = {250, 500}; // cm, known positions

Point trilaterate(float d1, float d2, float d3) {
  float x = (sq(d1) - sq(d2) + sq(anchor2.x)) / (2 * anchor2.x);
  float y = (sq(d1) - sq(d3) + sq(anchor3.x) + sq(anchor3.y) - 2*anchor3.x*x) / (2 * anchor3.y);
  return {x, y};
}

void loop() {
  float d1 = readUWBDistance(1); // DWM1000 ranging exchange per anchor
  float d2 = readUWBDistance(2);
  float d3 = readUWBDistance(3);
  Point pos = trilaterate(d1, d2, d3);

  float heading = readMagnetometerHeading(); // corrects drift from wheel odometry
  driveTowardWaypoint(pos, heading, targetWaypoint); // Mecanum path-follow + correction
}

Related Research

• Analysis and Accuracy Improvement of UWB-TDoA-Based Indoor Positioning System — PMC / Sensors
• GPS Accuracy — GPS.gov (official U.S. government source)