Australian Road Assessment Program (AusRAP)

The Australian Road Assessment Program (AusRAP) adopts iRAP’s methodology of assessing road infrastructure attributes. Up to 78 distinct road attributes for each 100-metre length of road and corresponding roadside environment are collected to form road attribute datasets for analysis.

A comprehensive list of all road attributes is contained in iRAP Methodology Fact Sheet 3: Road Attributes, available online at iRAP Methodology fact sheets - iRAP.

To assist with understanding the road attribute datasets at a high level, the AusRAP team have grouped these 78 attributes into 13 broad categories. These are detailed below.

Identifying Details

Identifying details provide context for the road being assessed, such as the road location and the source of the attribute information, including details of who has coded the data and when. These details also include the policy setting targets for the different road user types – in terms of star rating targets.

Carriageway Details

Carriageway details are key to understanding and evaluating the risk of a particular road environment. For example, a road with a single carriageway that carries traffic in both directions without median separation, is a largely different and relatively higher-risk road environment compared to a dual carriageway (separate carriageways with a sizeable distance between them) of multiple lanes in each direction.

Upgrade Costs

Three categories of cost (high, medium and low) are considered as attributes in the iRAP methodology. These options reflect the scale of costs associated with making improvements relative to the existing road environment, not the actual cost of implementing a treatment.

‘Low’ indicates that the road environment can be modified easily without costs like land acquisition or major earthworks. A ‘high’ cost indicates that there are likely to be acquisition costs, major service relocations and earthworks necessary for modifying the road environment to enable a treatment to be installed.

Vulnerable Road User Details

There are several attributes that relate to vulnerable road users (VRUs) (i.e., pedestrians, cyclists, motorcyclists). This includes attributes that identify the flow rate of VRUs (how many VRUs), providing insight into the level of VRU exposure in the road environment. This is complemented by attributes that identify VRU facilities (and their quality) that are in place.

Such facilities considered include footpaths and their separation from the roadway, crossing facilities, pedestrian fencing, on-road and separated cycle paths and facilities for motorised two-wheelers.

Collectively, the model considers both how many VRUs are present and the infrastructure standard in place to support those VRUs.

Land Use Type

Land use type provides context of the road environment for which specific treatments and/or focuses should be considered. In general terms, this reflects the need for different road safety infrastructure solutions between high VRU areas (such as around schools and universities) compared to rural farming and agricultural areas.

Speed (Limits and Operational)

Speed limits (and operating speeds), when complemented by effective enforcement activities, are one of the most impactful attributes to reducing road trauma. Speed impacts both the likelihood of a crash occurring and the severity of a crash if one does occur.  Furthermore, the relationship between speed and crash severity is generally exponential, meaning a small decrease in speed results in a much greater reduction in the number of crashes and their severity.

Barrier Types

Suitable and well-maintained barriers can contribute significantly to reducing road trauma. Median barriers are effective at mitigating head-on crashes from occurring and roadside barriers are effective at mitigating impacts with roadside risks (such as trees, bridge abutments or steep embankments). On routes with higher motorcycle traffic volumes, rub rails below the guard rail or wire rope can reduce the severity of impact for a motorcycle rider.

Delineation, Line Marking, Lighting and Sight Lines

The iRAP methodology considers a range of inputs that are commonly considered ‘low-cost’ and which can be cost-effectively implemented at scale (such as entire low-volume road corridors, where greater investment is not warranted based on overall risk).

This includes treatments such as delineation, line marking, lighting and the provision of adequate sight lines, which all help contribute towards the ‘readability’ of the road and help the road user to safely navigate the road or path that they are travelling on.

Roadside Hazards

Roadside hazards are considered as road attributes under two broad categories; the distance to the roadside hazard and the type of object that the hazard is. This helps identify the risks in the road environment by considering both the likelihood of impacting the roadside hazard and the severity of an impact with the hazard.

Pavement Specification and Condition

Pavement specification includes consideration of attributes like lane width, shoulder width, and rumble strips. These key factors reflect well-established road safety solutions where improvements to these attributes help reduce lane departure crashes. The iRAP methodology also considers the condition of the road and how the pavement condition may impact road users uniquely (this is especially relevant to bicycles and motorcycles, which may react more severely to some pavement conditions than 4 wheeled vehicles).

Intersection Specification and Accesses

The iRAP methodology considers various details about intersections including their type, channelization, volume, and quality. It also considers the number of property access points, to assess similar crashes (such as side-on impacts) that may occur at all access points to the roadway.

Curvature and Grade

Curvature is a well-established road safety consideration. Sharp bends, particularly unexpected corners, are more difficult for road users to safely navigate and often result in departures from the lane of travel, resulting in a head-on or road departure crash. Curve warning signs and Curve Alignment Markers (CAMs) can be used to highlight the safe cornering speed and alignment of the curve respectively. Often, roads need to be re-aligned to improve curvature and reduce the likelihood of crashes. Grade is also a similar consideration (although vertically) in road safety, albeit with less of a focus given the evidence-base and relative risk for curvature shows it is more commonly a significant road safety issue.

School Zone Warnings and Supervision

The iRAP methodology takes into consideration the infrastructure available at school zones, such as flashing lights and flags/beacons that help reduce the risk of vehicle-to-pedestrian conflicts. This also includes consideration of whether the location is supported by a school zone crossing supervisor.

Common Data Sources

Road authorities hold large amounts of data about their roads. This is necessary for them to be able to manage and maintain the road as a physical asset. Information such as the exact location, road alignment, posted speed limit, number of lanes, pavement depth, surface roughness and condition, location of intersections, to name a few, are already available in agency asset databases.

What is commonly missing from the asset data is roadside (non-road) information such as the condition of the road shoulder, roadside geometry, surrounding land use, trees and other fixed objects in the run-off-road area, vegetation growth that may impede a driver’s line of sight or be impacted in a crash, traffic average free flow speeds, and daily traffic volumes.

Traffic speed and volume information can be sourced from traffic surveys (completed by remote sensing – you may occasionally see two black rubber hoses stretched across a road used for counting traffic and measuring the speed of passing vehicles). These systems can also categorise if the vehicle is a car, truck, or motorcycle from the sensor information).

Traditional Gathering and Coding of AusRAP Data

Traditionally, common (asset) data sources (as mentioned above) are informed or complemented by a physical on-road survey conducted by fitting multiple video cameras to a survey vehicle. The video captured is linked geospatially via GPS data, to ensure there is an accurate record of the location and time of the survey data.

The video is then reviewed by a “coder”, who divides the road under observation into 100m lengths and identifies which road and roadside features, known as ‘attributes’, are present in each 100m length from the video. This attribute information is then ‘coded’ to enable AusRAP users to analyse the road data.

Manual collection and coding is relatively slow and labour intensive, which also carries the risk of human errors affecting data quality and additional time needed for quality assurance.

Fatal and serious injury crash data is also loaded into the analysis tool to generate supporting risk maps, showing collective and personal risk for a road user. The preferred data analysis tool can then be used to interrogate the AusRAP data to provide prioritised investment opportunities to improve road safety outcomes.

One benefit of the traditional approach to gathering and coding AusRAP data is that it can, in its simplest form, be completed with off-the-shelf camera technologies coupled with a GPS data capture system. This means risk-rating roads for targeted areas or localised projects doesn’t require significant technology or large-scale investment. Trained and accredited coders and analysts however, are still required to complete the data coding and analysis.

Recent Innovations and Future Advancements

More recently, traffic information has been obtained from “big data” sources.  Mobile phone and navigation system suppliers (such as Tom Tom, Sensis, Compass IOT, Ericsson, Apple, Google and many others) gather significant de-identified data from mobile devices carried in, or systems fitted to motor vehicles and can aggregate this to determine traffic volumes, speeds, congestion, harsh braking events (which may indicate issues like poor corner alignment or poor visibility) and much more.

iRAP introduced an initiative called AiRAP in 2019 to help improve access to, and application of, existing and emerging data sources globally, including advances in artificial intelligence, machine learning, vision systems, Light Detection and Ranging (LIDAR), telematics and other data sources.

AiRAP stands for the ‘accelerated and intelligent’ capture of road assessment program data using automatic, repeatable and scalable methods to support road safety assessment, crash risk mapping and investment prioritisation for all road users.

Under AiRAP, attempts have been made to utilise commercially available road imagery information such as Google streetview™ and Tom Tom’s Mobile Mapping (MoMa™) data. While informative, the fidelity of information available in Australia from such sources was insufficient to generate accurate road visualisations, meaning there remains a need to undertake a contemporary road attribute survey.

The latest AiRAP technique used for gathering high fidelity road data is multi head LIDAR scanners, coupled with colour video and GPS location sensors. The resulting LIDAR point cloud (example below) is rich in information and can be analysed using artificial intelligence to identify, extract, and code the desired road attributes, ready to be uploaded to iRAP's ViDA software or used with the local road authority's analysis tool of choice.

Image: LIDAR Point Cloud created view of section of Tasman Highway with overhead power cables. 3D view is simulated from point cloud captured from road level multi-head LIDAR scan. Source: Anditi.com (user portal)

The AI technology currently allows road data to be reviewed and coded at speeds of more than 400 km/h, which is over 125 times faster than traditional human video review techniques conducted by experienced coders, and without the risk of human error. A small number of queried items not resolved by the AI scanner can be visually checked against the video file, to make a manual determination. An Australian specialist spatial analytics company (Anditi) has been a global leader in the application of visualisation, classification and analysis of geospatial data for use in road safety applications, including AusRAP assessments. Anditi was the first company in the world to be accredited by iRAP for the conversion of LIDAR source data into iRAP road attribute data.

The use of LIDAR scans with AI interpretation makes the road assessments faster, more affordable and more reliable.

There are many other advantages to a road authority investing in survey quality LIDAR scanned point cloud of their roads and roadsides, in addition to completing a road assessment. Information such as bridge features, overhead power cable locations and their height, and tree branch height and thickness, when coupled with pre-existing road design information, can be used to identify over-size truck routes without having to leave the office.

Similarly, the point cloud can be used to determine the condition of side drains and roadside vegetation, to allow maintenance works to be scheduled, again without having to travel to and from the location.  The reduction in site visits by road asset managers increases productivity, reduces crash risk exposure while travelling for work, while also reducing travel related greenhouse gas emissions.

Suppliers like Anditi (mentioned above) can be accredited for the conversion of source data into iRAP attribute data. The AiRAP accreditation process will ensure data suppliers produce data in accordance with iRAP’s global standard format, regardless of the data source.

The accreditation process removes the need for complex data processing and storage for the data consumer. It also provides an understanding of the reliability of the data for different geographic regions, area types and road types, as well as when and how the data should be used. The process is flexible and accommodates attribute data derived from different types of source data, as well as different collection and processing methods.

For a list of iRAP accredited data suppliers please visit Accreditation - iRAP.

To read case study examples about AiRAP activities, see:

• AiRAP automation for Australian road safety: Final results - iRAP
• Anditi awarded largest ever AiRAP contract by Main Roads WA - iRAP
• Announcing another new ‘AiRAP’ accredited supplier – Agilysis Ltd. - iRAP

Star Rating Scores (SRS)

As detailed in the iRAP Methodology Fact Sheet 6, a Star Rating Score (SRS) is the score calculated for each 100 metre segment of road and for each of the four road users groups, using the following equation:

SRS = ∑ Crash Type Scores

where:

the SRS represents the relative risk of death and serious injury for an individual road user; and

Crash Type Scores =
Likelihood x Severity x Operating speed x External flow influence x Median traversability

where:

• likelihood refers to road attribute risk factors that account for the chance that a crash will be initiated
• severity refers to road attribute risk factors that account for the severity of a crash
• operating speed refers to factors that account for the degree to which risk changes with speed
• external flow influence factors account for the degree to which a person’s risk of being involved in a crash is a function of another person’s use of the road
• median traversability factors account for the potential that an errant vehicle will cross a median (only applies to vehicle occupants and motorcyclists run-off road and head-on crashes).

An SRS is only produced if a flow of the particular road user is recorded. For example, if no pedestrians are present, then no SRS is produced for pedestrians. SRS are also not produced when major road works are being undertaken.

Star Ratings

As detailed in the iRAP Methodology Fact Sheet 7, star ratings are determined by assigning Star Rating Scores (SRS) to the bands as shown in the table below. Separate bands are used for motorised road users (vehicle occupants and motorcyclists), bicyclists and pedestrians because their scores are calculated using different equations. That is;

• motorised road user scores are based on head-on, run-off road and intersection crashes
• pedestrian scores are based on walking along and across the road crashes
• bicyclist scores are based on riding along the road and intersections crashes.

Table A – Star Rating Scores and Corresponding Star Ratings

An Example of Star Rating Scores and Star Ratings Using the ViDA Demonstrator Tool
Using a generic ‘Low-standard Urban’ (with ‘Commercial access 1+’ property accesses) cross section.

Source: Star Rating Demonstrator (irap.org)

How Do the Stars Ratings Compare Practically?

There are clear differences between the level of safety on 1-Star roads compared to 3‑star roads and further, 5-star roads. The Global Status Report on Road Safety (WHO, 2018) provides a succinct high‑level summary of the key differences in road environment between these star ratings.

As presented in the chart below, there are clear differences between the road environments described by each star rating presented. These descriptions are not the only way to achieve the correlating star rating, rather a considered approach by road authorities can provide similar improvements in star rating. For example, a package of delineation, line marking, and improved sight lines accompanied by appropriate speed management can make a significant impact on the star rating of 1-star roads, towards a 3-star rating.

It is often beneficial to demonstrate changes in star rating scores alongside changes in star ratings as infrastructure investment plans can often have a timeframe of several years, or more than a decade. In such instances, variables such as traffic volume and road user behaviours (average or 85^th percentile speed) can change significantly meaning that star rating improvements may be impacted by non‑infrastructure variables.

Road safety treatments may result in a road moving from a star rating score of 22.4 (2-star) to 12.6 (2-star), and so a ‘star rating improvement’ to 3-star would not have been achieved however, as reflected by the change in star rating score, a significant safety improvement in the road environment has still been achieved.

Star ratings are not the only tool that should be utilised in the presentation and assessment of road safety risk. Other protocols, like Crash Risk Mapping, can also be used to provide further context. In such a case, it would be expected that Crash Risk Mapping would indicate a significant improvement in safety for the road given its improvement in star rating score.

Chart A – Star Rating of Roads – What Makes a Road Safe?

Source: Global Status Report on Road Safety 2018

Star Ratings and Injury Outcomes

The suggestion that higher star ratings results in lower levels of fatal and serious injuries was validated as part of a study conducted by McInerney and Fletcher (2013), on the Bruce Highway, Queensland, Australia. The findings, as presented in Chart A below, estimate that:

• crash costs on 2-star roads are 40% lower than on 1-star roads
• crash costs on 3-star roads are 61% lower than on 2-star roads
• crash costs on 4-star roads are 43% lower than on 3-star roads
• crash costs on 4-star roads are 86% lower than on 1-star roads.

As a general rule of thumb, for every improvement in star rating it is estimated that fatal and serious injuries will be reduced by approximately 50 per cent.

Chart B – The Relationship Between Star Ratings and the Cost of Fatal and Serious Injuries

From (iRAP_Star_Rating_and_Investment_Plan_Manual_English.pdf).

Source: https://irap.org/research-and-technical-papers/

3-Star Roads and Our Goal

It is not practical nor affordable to immediately upgrade every road to be a 5-star road. Incremental improvements in star rating score, whether that is within a star rating band or moving to a higher star rating band, are safety improvements.

Meaningful change, resulting in a significant reduction in fatal and serious injury crashes, will be achieved by reaching the national goal of at least 80% of travel on 3-star or better roads by 2030.

In most cases, more lives are saved by improving 1-star or 2-star roads carrying significant traffic volumes, than by improving 3-star or 4‑star roads carrying similar traffic volumes.

Using Results to Reduce Risk

Star rating maps and risk maps are produced to provide the user with visual information relating to how safe the road is (based on an assessment of road and roadside attributes) and what level of risk the road user is exposed to (based on historic, fatality and serious injury crash data).

These maps can be used to help select the safest route for a particular journey. Safer road choices are akin to choosing a safer vehicle. Both can help reduce the risks of a crash occurring or mitigate the consequences should one occur.

General Guidance – Interpreting Results

The information published on this website is intended as general guidance only and is provided to assist road users and others to understand road safety risks; to identify which roads, road segments or routes are safer or less safe than others and to modify their behaviour in ways that may reduce personal risk.

Despite this general information, it remains a fact that road crashes are somewhat random events in terms of when and where they may occur. Road crashes are also very chaotic in nature and the consequences of seemingly similar crashes may be significantly different because even the most minor variations in crash circumstances are amplified by the significant forces involved in the crash.

Risk is a function of both likelihood of occurrence and severity of outcome. Users of this website are advised that while the risk, or probability of an outcome may be increased or decreased by the absence or presence of different engineering treatments or by modifying personal behaviours, neither of these can provide an absolute guarantee that a crash will or will not occur, or a road user will or will not be killed or seriously injured on the road as a result of a crash.

Exclusion of Liability

The material published on this website is not intended to be relied upon as advice, and in particular the authors and publishers accept no responsibility for loss or injury suffered by any person as a consequence, direct or indirect, of anything contained in this website.

• Office of Road Safety
• Road Safety Sources (includes National Road Safety Strategy and NRS Action Plan)
• Bureau of Infrastructure and Transport Research Economics

• Australian Capital Territory
• New South Wales and NSW - Towards Zero
• Northern Territory
• Queensland
• South Australia
• Tasmania
• Victoria
• Western Australia

• Australasian College of Road Safety
• Department of Infrastructure, Transport, Regional Development, Communications and the Arts
• Infrastructure Australia
• Roads Australia
• National Transport Research Organisation (NTRO is the Australasian Centre of excellence for iRAP)