CrowdAtlas: self-updating maps for cloud and personal use
Proceeding of the 11th annual international conference on Mobile systems, applications, and services
| Hi-index | 0.00 |
Digital road maps have become essential to many aspects of our lives. Unfortunately, they have persistent quality issues, both in developing countries as well as in developed countries, evidenced by the recent Apple-Google map war. A survey of British drivers showed 26% have been directed by their GPS to go into no-entry areas, and the news periodically reports car accidents caused by or related to digital maps. In addition to correcting existing errors, maps need to be frequently updated to reflect the latest constructions, closures, and reconfigurations. TomTom estimates that roads change by as much as 15% each year. There are also growing demands for other types of maps, including off-road driving, cycling, hiking, and skiing maps. There are services for people to share GPS traces, but none creates navigable maps. Getting maps up to date and maintaining them involves a great deal of effort and delay. Today's maps are built by expensive geological surveys, supplemented by manual editing work from aerial imagery or corrections submitted by aggravated map users. NavTeq (now Nokia) employs more than 7,000 employees worldwide in its Location Content team to update maps. We present the CrowdAtlas system, which updates digital maps using the increasingly abundant GPS traces available as byproducts from a variety of sources: fleet management systems, telematics systems, and smartphone apps (e.g., navigation and location based services). We modified state-of-the-art map matching algorithms to accommodate the possibility that the existing map is incomplete. It uses the traces that match the map to monitor for road closures and fix road geometry. It uses tight clusters of trace segments from many vehicles that do not match the map in order to infer missing roads that connect to existing roads. The existing roads provide good segmentation of the traces to produce high quality clusters, enabling the automated (and even unsupervised) addition of missing roads. Using one week of traces from 70 taxis in Beijing, CrowdAtlas inferred 61km of new roads, which we uploaded to OpenStreetMap and became its first set of computer generated roads. To enable personalized maps, we also developed CrowdAtlas app based on OSMAnd, an open source navigation app. When acting as a GPS data source to CrowdAtlas server, it contributes data better optimized for map update with less communication. When acting in standalone-mode, it can add missing roads to its onboard navigation map. Instead of aggregating multiple GPS traces for high confidence, this app can add each new road immediately after the user traverses it, given user confirmation. We used our CrowdAtlas app in standalone-mode to map out major roads in a 4.5km^2 area of Shanghai Pudong in less than 30 minutes and build the walking map of the SJTU campus in less than a day. CrowdAtlas could fundamentally change the way people create and update maps. Together with incubating technologies that extract road metadata from street views and aerial imagery, modern cartography could be revolutionized, reducing or eliminating the expensive and slow manual mapping process used today. The high-definition version of our video presentation is available at vimeo.com/62912005, and our accompanying full-length paper describes the technical details of CrowdAtlas.