Open Source Robotics

A curated collection of high-quality open source robotics projects, tools, and software to propel the robotics community forward.

GNSS

  • Also called: global navigation satellite system, satellite navigation, satellite positioning
  • Category: State estimation & navigation › Localisation
  • Tags: localization, positioning, outdoor-navigation, sensor-fusion, field-robotics

Global Navigation Satellite System: constellations such as GPS, Galileo, GLONASS and BeiDou, and the receivers that use their signals to estimate global position, velocity and time.

Definition

GNSS, or Global Navigation Satellite System, is the general name for satellite constellations that broadcast timing signals for positioning, and for the receivers that use them. Four constellations have global coverage: the United States' GPS, Russia's GLONASS, the European Union's Galileo and China's BeiDou. Regional systems such as Japan's QZSS and India's NavIC add coverage over their areas. Modern receivers usually track several constellations at once, which gives more satellites in view and better geometry than any single system.

A receiver measures how long each satellite's signal took to arrive, giving a pseudorange that includes an unknown offset in the receiver's own clock. With at least four satellites it solves for three position coordinates and that clock offset. The result is a position in a global frame such as WGS 84, which a robot usually converts to a local frame such as east-north-up or UTM. Standalone accuracy is on the order of meters, limited by atmospheric delays, satellite orbit and clock errors, multipath reflections from buildings and terrain, and poor satellite geometry. Corrections improve this: satellite-based augmentation and differential GNSS remove errors shared by nearby receivers, and real-time kinematic (RTK) positioning uses carrier-phase measurements and a nearby base station or correction network to reach centimeter-level accuracy once the receiver has a fixed solution. Precise point positioning reaches high accuracy without a local base station but takes longer to converge.

For robots, GNSS is the main source of drift-free global position outdoors, but it has gaps. The signals are weak and are blocked or reflected by buildings, tree canopy, tunnels and roofs, and they can be jammed or spoofed. Update rates are low compared with inertial sensors, and a single antenna gives position and velocity but not orientation; heading comes from motion, from a second antenna, or from fusion with an IMU. In practice GNSS is fused with an IMU, wheel or visual odometry and sometimes lidar in a filter or factor graph. Combined GNSS/INS units perform this fusion internally and output a full pose, and vehicle datasets such as TartanDrive use GNSS-based pose as their reference state. Receivers also provide accurate time and a pulse-per-second signal that is widely used to synchronize sensors.

Why it matters

Outdoor mobile robots, drones, agricultural machines and autonomous vehicles need to know where they are on the earth, and GNSS is the only widely available sensor that provides this without drift. Its error sources and failure modes decide how far a robot can trust it and what has to back it up.

Key applications

  • Navigation of agricultural, construction and mining machines
  • Outdoor delivery and inspection robots
  • Drone flight control, waypoint missions and return-to-home
  • Reference trajectories and ground truth for datasets
  • Time synchronization of sensors with pulse-per-second signals
  • Georeferencing maps built with SLAM

In practice

An agricultural robot that drives along crop rows needs centimeter-level accuracy, so the team fits a multi-band RTK receiver and subscribes to a correction network over a cellular link. The receiver's position, its reported accuracy and its fix status go into an extended Kalman filter with the IMU and wheel odometry, so the robot keeps a smooth pose between GNSS updates and through short dropouts under trees. A dual-antenna setup provides heading even while the robot is stationary. The navigation stack converts latitude and longitude into a local east-north-up frame anchored at a corner of the field. When the receiver drops from a fixed RTK solution to a float or standalone solution, the filter increases the GNSS noise, and the robot slows down or pauses row following until the fix returns.

Common misconceptions

  • GNSS and GPS are the same thing — GPS is the United States' constellation, one of several GNSS systems alongside GLONASS, Galileo and BeiDou, and most modern receivers use several of them
  • GNSS gives centimeter accuracy — standalone receivers are accurate to meters; centimeter accuracy needs RTK or similar corrections and a clear view of the sky
  • A GNSS receiver tells the robot which way it is facing — a single antenna gives position and velocity only; orientation needs motion, a second antenna or an IMU
  • Height is as accurate as horizontal position — because all visible satellites are above the receiver, the vertical estimate is usually less accurate

Who should know this

  • Field and agricultural robotics engineers
  • Drone and autonomous vehicle developers
  • State estimation and sensor fusion engineers
  • Integrators of mapping and surveying robots

Related technologies

  • Real-Time Kinematic (RTK) Positioning
  • GNSS/INS Integration
  • Precise Point Positioning
  • Satellite-Based Augmentation Systems
  • NMEA 0183
  • UTM and ENU Coordinate Frames

Related terms

  • IMU
  • Sensor Fusion
  • Extended Kalman Filter
  • Odometry
  • Tightly Coupled Sensor Fusion
  • SLAM

Related projects

  • Cartographer: Google's real-time 2D and 3D SLAM system for multi-platform robotics
  • Nav2: ROS 2 navigation stack with path planning and obstacle avoidance for mobile robots
  • FAST-LIVO2: Fast, direct LiDAR-inertial-visual odometry from HKU-MARS, published in IEEE T-RO
  • GTSAM: Factor-graph optimization library powering many modern SLAM backends
  • AprilTag: The standard visual fiducial marker system for robot localization and pose estimation
  • GLIM: GPU-accelerated 3D lidar SLAM framework for any range sensor, from AIST

Videos

  • What is GPS/GNSS: Rami Tamimi | Professional Surveyor
  • How RTK works | Real-Time Kinematic for Precise GNSS Positioning: Emlid

This website requires JavaScript to function properly. Please enable JavaScript to explore our collection of open source robotics projects.