Mobile Device Acceleration Data Filtering for Road Quality Assessment

The patent extracts and removes human-carried device data from the acceleration data stream using heuristics that detect human movement patterns (walking, running) versus vehicle motion patterns. This extraction allows the system to maintain road quality assessment capability while eliminating the contaminating effect of human-carried data, thus resolving the contradiction between reliability improvement and measurement precision deterioration.

The patent introduces an intermediary processing layer that analyzes acceleration data through multiple heuristics (speed, endurance, frequency, acceleration metrics) to distinguish between human-carried and vehicle-based data. This intermediary analysis layer enables the system to maintain both road quality assessment capability and measurement precision by filtering out contaminated data before it affects the assessment results.

The patent segments the data processing task into multiple independent heuristics (speed heuristic, endurance heuristic, frequency heuristic, acceleration heuristic), each handling a specific aspect of the analysis. This segmentation allows the complex filtering task to be broken down into manageable components that can be processed in parallel, improving measurement precision while keeping the overall device complexity manageable through modular architecture.

The patent applies multiple heuristics simultaneously, using more analysis than the minimum required (excessive action). By applying speed, endurance, frequency, and acceleration heuristics together, the system ensures thorough filtering of human-carried data, improving measurement precision. The redundancy in this approach provides robustness against any single heuristic failing, justifying the increased complexity through improved reliability.