Applying a ToF/IMU-Based Multi-Sensor Fusion Architecture in Pedestrian Indoor Navigation Methods

Area of Science:

• Robotics
• Navigation Systems
• Sensor Fusion

Background:

• Low-cost Inertial Measurement Units (IMUs) in indoor Inertial Navigation Systems (INS) suffer from unreliability and inaccuracies.
• Sources of error include imperfect modeling, sensor fusion, heading drift, IMU biases, and calibration issues.

Purpose of the Study:

• To address the unreliability of Micro-Electro-Mechanical System (MEMS)-based pedestrian INS.
• To develop a novel multi-sensor fusion method for improved indoor pedestrian localization.

Main Methods:

• A novel fusion method using a Time of Flight (ToF) sensor and dual chest/foot-mounted IMUs with online calibration.
• An Extended Kalman Filter (EKF) for estimating attitude, position, velocity errors, and IMU biases.
• A fusion architecture integrating ToF and foot-mounted IMU for velocity, with chest-mounted IMU for attitude and a corridor detection filter for heading drift.

Main Results:

• The developed system demonstrated promising and resilient results in 2D corridor spaces for up to 11 minutes.
• Achieved a position Root Mean Square (RMS) error of less than 3 meters.
• Attained a final-point error of less than 5 meters.

Conclusions:

• The novel multi-sensor fusion approach significantly enhances the reliability and accuracy of low-cost MEMS-based pedestrian INS.
• The system effectively mitigates common error sources, offering robust performance in indoor environments like corridors.