PALMS: Plane-based Accessible Indoor Localization Using Mobile Smartphones

Area of Science:

• Robotics
• Computer Science
• Geographic Information Systems

Background:

• Indoor localization is challenging due to GPS limitations.
• Existing methods often require extensive environmental mapping or constant visual input.
• Mobile smartphones are ubiquitous, offering potential for indoor navigation solutions.

Purpose of the Study:

• To develop an innovative indoor global localization and relocalization system for smartphones.
• To improve localization accuracy and reduce convergence time compared to traditional methods.
• To create a scalable and practical indoor navigation solution without prior environmental fingerprinting.

Main Methods:

• Utilizing publicly available floor plans for localization.
• Adopting a dynamic localization approach combining instantaneous observation and odometry data.
• Introducing a particle filter initialization method with the Certainly Empty Space (CES) constraint and principal orientation matching.

Main Results:

• PALMS creates a spatial probability distribution for improved device location accuracy.
• The system significantly reduces particle filter convergence time.
• Experimental evaluations show PALMS outperforms traditional uniformly initialized particle filters.

Conclusions:

• PALMS offers a more efficient and accessible approach to indoor wayfinding.
• The system eliminates the need for prior environmental fingerprinting, enabling scalability.
• This method provides a practical solution for smartphone-based indoor navigation.