Causes of crashes affecting pedestrians Michael Nieuwesteeg

Presentation outline Background – pedestrian trauma Interviewing injured pedestrians What causes pedestrian crashes Tackling pedestrian injuries New TAC programs for pedestrian and cycling safety

Pedestrian trauma Pedestrians # (%) Deaths per year 48 (16%) # (%) Deaths per year 48 (16%) Hospital >14 days claims 180 (20%) Hospital claims 530 (12%) TAC costs $65m (10%)

Pedestrian trauma Pedestrians comprise 11.3% of all serious casualties Pedestrians in 40, 50 and 60 speed zones make up 8.8% of all serious casualties 60 zones intersections are particularly problematic Almost even split between intersection and mid-block

Pedestrian trauma

TAC research with injured pedestrians TAC commissioned exploratory survey research with its clients (persons compensated for road transport injury) who were injured as pedestrians in: 16-39 or 60+ age groups 40, 50 or 60km/h zones The research sought to identify: typical crash circumstances risk factors in pedestrian crashes issues that warrant further exploration In recognition of the increasing share of pedestrian trauma in TAC statistics, the TAC’s road safety team sought to gain insight into the pedestrian crash problem. We saw this particular research as a first, exploratory step in this direction. To restrict the scope of the research, we determined to focus on those cohorts that had the biggest impact on the TAC: the elderly (who have the costliest claims) and adults aged 16-39 (who are most likely to be intoxicated). These groups present substantial challenges to road safety organisations. Research aims were kept simple and the sample size kept small, as we saw this as exploratory research. Our aims and sample frame could be refined in future waves of this research. The TAC is fortunately placed in that it has access to persons injured in transport accidents. The TAC’s clients (ie. persons compensated for injuries sustained in transport accidents) already have an established relationship with the TAC through their claim manager, and have often been contacted for internal research purposes with a focus on client satisfaction. In 2008 I was involved in the first TAC attempt at surveying its clients for road safety purposes. This was a research project that invovled surveying 500 TAC clients injured while motorcycling. The success of that project has encouraged us since, leading to the research project I am sharing with you today. Pedestrian crashes are generally poorly understood for obvious reasons: police reports of crashes are usually compiled after the injured pedestrian has been whisked off to hospital. The police will have very little evidence available other than a statement from the vehicle driver, and possibly from an eyewitness. Hence our interest in studying crash circumstances.

TAC research findings Crash types by age of injured pedestrian % of crashes Pedestrians aged 16-39 Pedestrians aged 60+ Number of crashes (n) Pedestrian crossing road 75% 78 71% 71 79% Crossing intersection 40% 38 35% 41 46% Crossing mid-block 31% 37 34% 24 27% Crossing roundabout 5% 3 3% 6 7% Pedestrian not crossing road 26% 32 29% 19 21% Car off-road / out of control 10 9% 4 4% Standing/Walking on road 7 6% 1 1% Private/Friendly Car-park Driveway cross-over footpath 2 2% Struck while boarding/alighting Fall while boarding/alighting Other

TAC research findings Intersection crashes 58% involved vehicle turning right, usually exiting, ped usually on far side of road (completing crossing) Driver usually at fault In many of these crashes where the driver was at fault there were elements present that made visibility difficult (low light, dark clothing, rain) Limited understanding of impact of traffic lights. Red-light running appears to be uncommon Band-aids

TAC research findings Mid-block crashes 70% involved pedestrian being hit on near side 20% involved reversing vehicles In 64% of cases an intersection was within 50 metres In a third of cases the vehicle had just pulled out of a parking spot or turned into road 22% involved pedestrians who were crossing at a signalised pedestrian crossing or zebra crossing

TAC research findings Other factors A quarter of younger group had consumed alcohol pre-crash Pedestrians usually injured while on routine trips in familiar locations Distraction, which was an issue for the younger group, often co-occurred with impairing factors and mid-block crashes Difficult light or weather conditions seem to contribute more to the intersection than mid-block crashes, probably by adding another level of difficulty to an already complex driving task

TAC research findings Age differences Younger group significantly more likely to be at fault, to have consumed alcohol, to be distracted, stressed, tired Younger group more likely to be injured in poor light conditions Older pedestrians less likely to cross mid-block

Roundabout and other crashes TAC research findings Issue Younger Older Intersection Mid-block Roundabout and other crashes Pedestrian at fault 34% 12% 16% 43% 18% Familiar with location 90% 89% 91% 93% 83% Heavy or congested traffic 23% 10% 11% 19% 26% Poor light conditions 46% 22% 44% Raining 8% 14% 5% Tired/fatigued 15% 3% 6% 9% Stressed/anxious 2% 7% Distracted 24% Impaired by alcohol 25% Impaired by drugs 0% NA Impaired mobility Summary of risk factors Older peds are unlikely to be at fault, distracted or impaired, and impaired mobility is not a large problem in terms of the overall pedestrian injury issue. Younger peds are more likely to be involved in crashes in which we find more complex conditions

A system view Limits of human performance - functionality - biomechanically in collisions Limits of vehicle performance - crash avoidance - crashworthiness Physical design of roads and roadsides Vital role of speed TAC believes almost all crashes can be avoided by creating a system that recognises and caters for the limits of human and vehicle performance and the role of speed in crash causation and injury severity. The most dependable method to achieve this is through the design of roads and roadsides.

Modified Behavioural Sequence Model Function/event sequence Search Detection Evaluation Decision Human action Vehicle/system action Response Predisposing factors Driver factors Ped/cyclist factors Vehicle factors Environment factors Post-crash factors Safe System Pedestrian/cyclist crash causation is complex and multifactorial. Fortunately, similarities and discrepancies among such crashes are such that a typology of pedestrian/cyclist crash causation can be developed. This is a promising finding because it posits the notion that a finite number of factors or combinations can be found that sufficiently explain the cause of a majority of pedestrian/cyclist crashes. In 1971, Snyder and Knoblauch introduced the Behavioural Sequence Model, which includes two main features: The function/event sequence: dealing with the sequence that leads to the pedestrian crash; and Influencing/predisposing factors: dealing with factors which influence the behaviour within the sequence. The function/event sequence feature attempts to construct general, but not simplistic, sequence of events that applies to all pedestrian/cyclist crashes. While the specific events may vary for each crash, certain functions are performed by the participant road users that may result in the events leading to a crash, or prevent a contingent collision. These basic functions are Search: referring to the focus of the driver's and pedestrian’s/cyclist’s attention and their behaviour that influence their perception of the environment. Detection: referring to their actual perception of the environment. Successful detection means that they are aware of the other party. Evaluation: referring to their evaluation of what they have perceived (the other party and their surroundings); if "successful," results in recognition of the threat of a collision and the need for action to avoid it. Decision: referring to the determination of the action necessary to avoid a collision. Human action: referring to the motor behaviour of the driver and pedestrian to implement the avoidance decision. Vehicle action: referring to the response of the vehicle to the motor behaviour of the driver. Snyder and Knoblauch proposed four main class of predisposing factors that influence the function/events in the behaviour sequence. They argue that to understand the cause of pedestrian crashes, the specific variables in these groups that influence the function/event sequence must be known. The main classes are Driver factors Pedestrian/cyclist factors Vehicle factors Environmental factors (includes physical environment and speed regime)

When the system fails…

When the system fails… Sequence Hypotheses System Support Intervene Seek Ped: maybe Driver: did (?) Signal timing (haste) Design (distraction?) Detect Ped: maybe? Driver: didn’t Raised crossing; Lighting; Vehicle design/technology Evaluate Ped: maybe (error, mishaps?) Vehicle technology Decision Ped: maybe Action Ped: maybe – non-preventive Vehicle technology (Autonomous Emergency Braking) Response NA

When the system works… Safe System travel speeds Smart roads (ped detection) Vehicle technology (ISA, AEB) Road design, e.g., low speed environment, raised crosswalks Responsible road use

When the system works

TAC supporting safety infrastructure LGA grants $100m pedestrian and cycling safety fund Small scale infrastructure for cyclists and pedestrians Safe System focused $1 million per annum $:$ up to $100,000 for infrastructure treatments or Up to $25,000 grant to develop a project Area wide or route based preferred

LGA grants Round 1 – August 2014 Round 2 closes mid August 2015 97 applications, 27 funded 14 Metro Melbourne 13 Regional Victoria (2 in large cities) 11 focus on pedestrians 10 focus on cyclists and pedestrians 6 focus on cyclists Many include speed reduction to 40 km/h Round 2 closes mid August 2015 Small scale infrastructure for cyclists and pedestrians Safe System focused $1 million per annum $:$ up to $100,000 for infrastructure treatments or Up to $25,000 grant to develop a project Area wide or route based preferred