Skip to content

Software Requirements Odometry System

Simran Cheema edited this page Jul 15, 2026 · 1 revision

3.4 Odometry System

3.4.1 Objectives

  • Fuse odometry, IMU, GPS, and ZED VO for accuracy.
  • Provide local and global pose via EKF.
  • Visualize in Mapviz and diagnostics tools.
  • Remain reliable under partial sensor loss.

3.4.2 Functional Requirements

FR-4.1 Sensor Data Integration

  • Encoders → baseline odom.
  • IMU reduces short-term drift.
  • GPS corrects long-term drift.
  • ZED2i VO as additional source.
  • All data time-synchronized via headers.

FR-4.2 State Estimation (EKF Fusion)

  • FR-4.2.1 Local EKF fuses encoders + IMU → /localization/odom_local (odom frame).
  • FR-4.2.2 Global EKF fuses local EKF + GPS + ZED VO → /localization/odom_global (map frame) and publishes map↔odom.
  • FR-4.2.3 navsat_transform consumes local orientation (from local EKF/IMU) + GPS → /localization/navsat/odometry (world frame).
  • FR-4.2.4 Publish standard /odom, /tf for other subsystems.
  • FR-4.2.5 Graceful degradation under GPS loss (rely on local EKF).
  • FR-4.2.6 Parameters in ekf_local.yaml, ekf_global.yaml.

FR-4.3 Visualization & Base Station Feedback

  • Mapviz shows trajectory/position.
  • Display local (odom_local) and global (odom_global) paths simultaneously.
  • UI shows drift and sensor health.
  • Logs include timestamped odom, EKF outputs, drift metrics (CSV export).

FR-4.4 Simulation & Testing

  • Testable in RViz/Gazebo with simulated encoders/IMU/GPS/ZED.
  • 100 m straight-line drift tests (sim + field) with bag/CSV.
  • Validate EKF vs ground truth (e.g., GPS path).
  • Automated scripts compute error, drift %, update rate.

3.4.3 Non-Functional Requirements

ID Type Requirement
NFR-4.1 Accuracy Drift < 5% over 100 m under normal conditions.
NFR-4.2 Performance Updates ≥ 30 Hz.
NFR-4.3 Reliability Stable estimates during temporary sensor dropouts.
NFR-4.4 Robustness Handle failures (continue with IMU+encoders when GPS absent).
NFR-4.5 Testability Configurable logging for inputs and EKF outputs.

3.4.4 Verification Criteria

  • Local EKF drift < 2% (short-term sim + field).
  • Global EKF integrates GPS + ZED VO for long-term accuracy.
  • Total drift < 5% over 100 m straight-line.
  • Mapviz shows local/global estimates at ≥ 30 Hz.
  • Sim confirms EKF performance/parameter stability pre-hardware.

Clone this wiki locally