Authors: Huihai Wang*,
Topics: Geographic Information Science and Systems
Keywords: Mobile Sensor, Sidewalk Anomaly, IOS system, Android system
Session Type: Guided Poster
Presentation File: No File Uploaded
Sidewalk plays an important role in sustainable transportation systems. Bad conditions of sidewalks will bring risks to pedestrians. According to the Americans with Disabilities Act of 1990, sidewalks must be ‘accessible to and usable by’ people with disabilities. However, traditional methods for road condition monitoring are labor-intensive and expensive. Therefore, this research aims to find an efficient way to detect sidewalk anomalies on a large scale. An algorithm is developed to process acceleration data collected by smartphone GPS and acceleration sensors. This algorithm contains three major steps: reorientation of acceleration data; robust peak detection; noise removal and anomaly detection. Experiments using smartphones with different operations systems (IOS system versus Android system), as well as at different speeds were conducted (5m/s and 2m/s), and their performances were compared. The result shows that the highest accuracy of IOS system is about 90% when speed is 5m/s, which is much better than that of Android system at the same speed. Finally, a sidewalk anomaly map can be created by using smartphone with IOS system amounted on a scooter at a speed of 5m/s. According to our analysis, three conclusions can be made. First, an efficient and automated method was developed to detect sidewalk anomaly by using in-built sensors of smartphones. Second, operation system and riding speed can influence the detection accuracy. The best accuracy was derived with an IOS system at an acceptably high speed. This efficient method could be helpful for planners as well as government officers for sidewalk condition evaluation.