The gap between cycling practices and mapping services for smart cycling

Abstract

Cycling is gaining priority in urban mobility policies and smart cities should be prepared to offer smart cycling services that help to promote such transition. A key element for many of those services is a Geographic Information System and particularly a road network model that represents all the possible roads in a city. The main contribution of this work is the identification and characterization of an important gap between the GPS traces corresponding to the real routes made by cyclists and their representation in the road network model of Open Street Map (OSM). More specifically, there are parts of the GPS traces that cannot be mapped into the road network of OSM, because they have no matching representation. We call these segments, ad-hoc segments. To develop a deeper understanding about this problem, we collected data from a specially designed route and analysed how the respective trace was mapped into OSM. We identified all the occurrences of adhoc segments and categorized them according to their root cause. The results suggest that the main overall cause of these problems is directly linked with limitations of the road network model, such as missing roads. This seems to indicate that road network models, as commonly used in OSM, do not properly address the specificities of cycling, or micro-mobility in general. This problem has a major impact in the quality of the services provided to urban cyclists and in the value of the data available for micromobility planning and management.

High-Performance C60 Coupled Ferroelectric Enhanced MoS2 Nonvolatile Memory

“…Computers with the traditional von Neumann architecture have encountered bottlenecks in handling the explosive growth of data in the information age as its separated processing and storage units, and the data transfer size inside a computer is limited by the bandwidth . In addition, the serial execution approach between the memory cell and the processor was high time and energy-consuming. , Therefore, the new computing-in-memory (CIM) system based on nonvolatile memory (NVM) devices demonstrated a great prospect because it can not only store program instructions in situ according to the programming operation but also output the data in parallel to complete the calculation . However, with the advent of the post-Moore era, NVM devices based on conventional semiconductors have been plagued by scaling-down limitations in large-scale integration and high-performance preparation.…”