Spatial Trends of London’s Road Accidents

on YouTube ↗15m

" summary

Uses DfT STATS19 point data for 2019, aggregates into H3 hexagons (0.75 km²), normalizes counts and runs local Moran’s I spatial autocorrelation to map hotspots. Finds a dense central high‑high core, surprising outer high‑low pockets, an early‑morning crash peak, and fewer accidents on high‑precipitation days. Single‑year results fail to be explained by DfT attributes alone, so next steps bring in demographics and infrastructure (collaboration with CARTO) to test causal drivers.

AI-written summary · about

" chapters (8)

1. 0:00Introduction and background
2. 1:13Context and motivation
3. 3:40Data overview and preparation
4. 5:49Temporal trends in accidents
5. 6:55Spatial autocorrelation analysis
6. 8:43Hotspot characteristics and findings
7. 10:30Next steps and future work
8. 11:35Q&A session

" related talks

• 01h 03mUnderstanding and Shaping Urban Mobility Using Dollars and Data IIBalaji Prabhakar · Real-Time Decision Making Boot Camp, Spring 2018
• 10mCivic Hacking & Urban Informatics: Mapping Data with GeoPandasKatelyn Morrison · PyOhio 2020
• 37mLocation Analytics For Smart Grid ReliabilityVivian Sultan · Data Con LA 2019
• 22mR & Space : GI Science with RHadrien Commenges · satRdays Paris 2019
• 01h 15mSunday KeynotesRon Galperin, Brian Kursar, Larry Dominique, Lacey Plache, Brian Dolan, Michael Berman, Bryan Kirschner · Data Con LA 2020
• 32mData&AI in TomTom: How are are transforming mobilityPierLuigi Cassale · Data Science Conference 4.0
• 32mUsing spatial data to model biodiversityJane Elith · FOSS4G SotM Oceania 2018
• 30mLocation Intelligence and Spatial Data Science: from geographic coordinates to business insightsPiergiacomo Fonseca · PyCon Italia 2023