Loose Coupling of Wearable-Based INSs with Automatic Heading Evaluation.

Area of Science:

• Robotics
• Sensor Fusion
• Human-Computer Interaction

Background:

• Pedestrian position tracking using inertial sensors is a key research area.
• Existing inertial navigation systems (INSs) typically rely on a single inertial measurement unit (IMU).
• The future of wearable devices necessitates advanced multi-IMU tracking solutions.

Purpose of the Study:

• To implement and evaluate an INS utilizing two wearable IMUs.
• To develop a loosely coupled fusion approach for combining data from pocket-mounted and foot-mounted IMUs.
• To introduce a novel 'smart update' method for optimizing INS performance.

Main Methods:

• A loosely coupled fusion strategy was employed to integrate outputs from two wearable INSs.
• The system utilized a pocket-mounted IMU and a foot-mounted IMU.
• A novel concept for assessing heading estimation quality, subject to patent, was developed to enable the 'smart update' functionality.

Main Results:

• The proposed 'smart update' fusion method intelligently combines INS outputs, prioritizing the least erroneous data.
• A new method for dynamically evaluating heading estimation quality was successfully implemented.
• The loosely coupled fusion demonstrated a 10 cm/s improvement in position error rate compared to individual foot or pocket INS in 95% of cases.

Conclusions:

• The 'smart update' fusion of dual wearable IMUs offers a significant advancement in pedestrian position tracking.
• The novel heading quality estimation method is crucial for the success of the smart update strategy.
• This approach enhances INS accuracy and reliability for future wearable technology applications.