Introducing 

Prezi AI.

Your new presentation assistant.

Refine, enhance, and tailor your content, source relevant images, and edit visuals quicker than ever before.

Loading content…
Loading…
Transcript
  • Traffic Counts & Casualty data
  • All 'open'
  • Filtered for London

Casualties

  • Cyclists only

  • Severity Breakdown (Slight, Serious, Fatal)

  • 18,000 records 2011-14

  • Definite potential to help cyclists & influence policy

  • Sharing / Assurance

  • Other data sources (e.g. Congestion, Google, Strava)

  • Time of Day

  • Weather

  • Clustering

  • Predictive Modelling / Machine Learning

  • Apply techniques to other streams of DfT work (e.g. Traffic Flows, Road Incidents, Rail etc.)

London Enhanced Cycling Traffic Counts 2014

London Enhanced Motor Vehicle Traffic Counts 2014

London Motor Vehicle Traffic Counts 2014

London Cycling Casualties 2011 - 14

Assigning Traffic Counts to Casualty Points:

1. If casualty is on a named road (e.g. A14) the two nearest count points on that road are identified.

2. If the road is not named (e.g. a generic ‘C’ road) or no count points exist for a named road, nearby roads of the same type are queried.

3. Once the most appropriate two count points are identified, counts are assigned based upon the relative distances away.

4. Once all counts are assigned, the derived count data is concatenated to the actual count data and a new Raster is produced for evaluation.

5. Included ‘Slight’ casualties to give more data.

Nearest Neighbour Algorithm

Example 1

Example 2

Exploring the data part 2

Exploring the data part 1

Tom Ewing

Improving Cycle Safety with Data

Data

Next Steps

Cycle Safety Project

Traffic Counts

  • Road (e.g. A19)

  • Road Type (A, B, etc.)

  • Vehicle Type

  • 2100 Count Points

  • Distinguish 'risky' from 'busy'

  • Help cyclists plan safer routes

  • Inform policy & network improvements

Tools

  • Python

  • Quantum Geographic Information System (QGIS)

  • Google Earth

Learn more about creating dynamic, engaging presentations with Prezi