Intelligence-led safe road systems
Advanced algorithms transform our strategic road networks – saving lives
Road traffic collisions and congestion cost the UK more than £72 billion a year. Latest Department for Transport figures showed that 1,580 people died on our roads (year to June 2020), and drivers can spend an average of 31 hours a year in rush hour queues.
Highways England (HE) was tasked by the Government, in 2015, with reducing the number of deaths and serious casualties on the UK’s 4,300-mile strategic road network (SRN) by 40% before the end of 2020. To achieve this target, HE needed accurate, high-quality data and modelling.
We have developed artificial intelligence (AI) based collision mapping and risk modelling that provide an unprecedented accuracy of 99% in pinpointing accident hotspots.
Since 2014, in partnership with AECOM, we have been helping HE to deploy these algorithms and develop a holistic understanding of road safety – looking at how vehicles, people and the design of road infrastructure interact.
This has supported targeted investment in safety measures and the successful implementation of a new range of interventions spanning road engineering, behavioural change campaigns, and technology improvements to vehicle maintenance.
Accurate accident data improves road safety
- Our AI-based algorithms provide HE with vastly improved collision data that is 99% accurate.
- HE has used this data to develop a new approach to road safety – the Safe System Model (SSM).
- The SSM has enabled HE to meet its road safety improvement target.
- Our research has facilitated the nationwide mapping of collision risk, identifying high-risk routes and vulnerable road users such as motorcyclists.
- Identified collision hotspots and targeted interventions reduced casualties across the entire SRN by 20% in just four years (2014-18) – despite an 8% increase in total miles travelled.
Driving Highways England’s safety strategies
- Our evidence underpins Highways England’s Road Safety Delivery Programmes.
- The successful reduction in collisions led to the Treasury’s unique decision, in 2019, to invest all vehicle excise duty collected in England (£27.4 billion) in future road improvements.
Indeterminate locational accuracy makes effective analyse of geo-spatial data extremely challenging.
In 2006, we began to address this issue. Our innovative and transferable statistical and artificial intelligence-based map-matching algorithms make analysis of these data far more reliable – achieving over 99% accuracy so that effective road safety measures can be devised and implemented.
In 2018, we used the algorithms to analyse data, supplied by the Department for Transport, comprising more than 70,000 traffic collisions from the preceding six years. This analysis allowed us to create more than 500 spatio-temporal safety risk maps, identifying fatal hotspots and high-risk routes.
These maps have enabled Highways England – in partnership with AECOM – to identify the factors underlying the frequency and severity of traffic collisions, and to develop and evaluate effective collision prevention schemes.
Our work with Highways England to improve UK road safety is ongoing.
- Highways England
- Highways England