United States Department of Transportation
i
Bus on the Edge: Continuous Monitoring of Traffic and Infrastructure – Year 2
-
2021-08-31
[PDF-4.01 MB]
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Resource Type:
-
Geographical Coverage:
-
Edition:Final Report
-
Corporate Publisher:
-
Abstract:For the purpose of safety and liability, transit buses nowadays usually have cameras installed to observe the environment around the buses, together with some other sensors like GPS and IMU. Such data sources are undoubtedly valuable to the ambitions of a resilient and intelligent transportation system. For example, these live data streams can be analyzed and used for abnormal event detection, traffic modeling and infrastructure monitoring and thereby provide input for up-to-date detailed maps of roads and traffic. Up-to-date detailed maps are one essential component of autonomous driving, but they are also needed for traffic management, planning, and infrastructure maintenance. Moreover, the mobility of transit buses also makes the freshly-captured visual data the favorable input for executing ad hoc search queries. Example applications include searching for a missing child or lost items along the roads as well as collecting some distilled images to construct a dataset for research given a specific target. Other use cases include traffic counts, counting of parked cars, observations of road construction, pothole detection, detecting landslide precursors, measuring snow cover, or observing crossing of wildlife. However, it is a difficult task to process and analyze the live bus data. For one thing, it is impractical to send all these data to the cloud for analysis because of the bandwidth and storage constraints. In addition, the real-world collected data are extremely redundant despite their application value. Only a small fraction of the data has the information we are interested in. This is exactly where edge computing can play an important role. Specifically, we will have an in-vehicle computer installed on the bus to have preliminary processing of the raw input in real time and only send the frames of interest to the nearest cloudlet server for further analysis. In this context, we develop an efficient data collection and analytics platform called BusEdge, the goal of which is to tap into the valuable but redundant bus data to achieve data refinement and analysis in an efficient manner.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type:
Related Documents
-
[PDF - 1.64 MB]
[PDF - 977.86 KB]
[PDF - 1.59 MB]
