The first road crossing was invented in the 1930’s. Since then, city life has changed. But pedestrian crossings haven’t kept up. 7,000 incidents happen on Britain’s crossings every year*.
What if we could rethink how a road works, and design a crossing that can see what’s happening on it, and adapt itself to help stop accidents from happening at all?
Could a smarter crossing save lives?
See the Partners section below for more information about the technology used to develop The Smart Crossing.
*Based on data from Road Safety Analysis who recorded 29,028 pedestrian casualties injured on or near a pedestrian crossing in Great Britain between 2011 and 2015
Since the first road crossing was invented, city life has changed. In 2012, the decline in road accidents plateaued, leaving experts wondering if a better solution, could be a better crossing.
Umbrellium designs and builds technological tools to support citizen empowerment and high-impact engagement in cities.
They joined the project as lead technology partner, with responsibility for the design and build of The Smart Crossing.
Transport Research Lab
TRL enable world-class transport and mobility solutions that underpin tomorrow’s economy and society.
We worked with TRL to identify opportunities to pioneer a safer way to cross by looking at accident research and city infrastructure.
University College London
UCL is a public research university based in London.
We worked with transport, road safety and design experts from UCL to understand the best way to modernise road markings, and animate them in ways pedestrians, drivers and cyclists will instinctively understand.
The base programming codes used to develop The Smart Crossing are:
ofxBlackMagic: Simplified and optimized Black Magic DeckLink SDK grabber
ofxCv: ofxCv represents an alternative approach to wrapping OpenCV for openFrameworks.
OpenCV: OpenCV (Open Source Computer Vision Library) is an open source computer vision and machine learning software library.
ofxDarknet: ofxDarknet is a openFrameworks wrapper for darknet.
Darknet is an open source neural network framework written in C and CUDA. It is fast, easy to install, and supports CPU and GPU computation.