Probabilistic Motion Maps for Indoor Device Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Indoor navigation and localization of mobile devices is challenging due to unreliable satellite-based systems and inconsistent behavior of sensors in varying indoor environments, leading to inaccurate trajectory determination and positioning errors.
Innovation Solution
A probabilistic approach representing indoor spaces as a probability map with motion transition probabilities at each point, using sensor data and machine learning to predict and correct device motion, employing a hierarchical tessellated grid for scalable representation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If satellite-based navigation systems are used for indoor positioning, then outdoor positioning accuracy is maintained, but positioning reliability deteriorates in enclosed spaces due to line-of-sight requirements
Solution Approach 1:
The patent introduces an intermediary probabilistic motion transition model that bridges satellite-based positioning and sensor-based positioning. This model uses probability distributions to represent motion transitions between locations, allowing the system to maintain positioning reliability in indoor environments by transitioning from direct satellite positioning to probabilistic motion prediction when satellite signals are unavailable.
Solution Approach 2:
The system dynamically switches between different positioning modes based on environmental conditions. When satellite signals are available, it uses GPS positioning; when unavailable, it transitions to sensor-based positioning with probabilistic motion transition models. This dynamic adaptation resolves the contradiction by maintaining reliability across different operating conditions.
2Reliability
If sensor data is used for indoor positioning, then positioning capability is maintained in enclosed spaces, but measurement precision deteriorates due to inconsistent sensor behavior in varying environments
Solution Approach 1:
The patent transforms deterministic sensor measurements into probabilistic representations. Instead of relying on fixed sensor readings that vary inconsistently in different environments, the system models motion transitions as probability distributions, changing the parameter representation from deterministic to probabilistic to account for environmental variations and improve measurement precision.
Solution Approach 2:
The system performs preliminary actions by collecting and analyzing motion data to build probabilistic motion transition models before actual positioning is needed. These pre-established probability distributions capture typical motion patterns in specific indoor environments, enabling more accurate positioning when sensors are used during operation.
3Device complexity
If traditional deterministic motion models are used, then computational simplicity is maintained, but prediction accuracy deteriorates in complex indoor environments
Solution Approach 1:
The patent changes the parameter representation from deterministic values to probabilistic distributions. This allows the model to capture the uncertainty and variability inherent in human motion within indoor environments, improving prediction accuracy while maintaining computational tractability through efficient probability manipulation.
Data Source
AI summary
A method and a device for determining motion of a device are described. In an example, a plurality of points-of-interest (POIs) are identified in a pedestrian area of an indoor space and, with each POI, a probabilistic representation of motion transition of a device from the POI is associated. The probabilistic representation of motion transition is indicative of a probability distribution of likely directions of motion that the device can exhibit from the POI. Further, a map of the indoor space is generated based on the association and the map is usable for ascertaining motion of the device in the indoor space.