Transportation Safety and Human Factors

Introduction and Context

Transportation safety is the systematic prevention of death, injury, and property loss arising from the interaction of road users, vehicles, infrastructure, operating environments, and institutions. It is not limited to identifying hazardous locations after crashes have occurred. Contemporary practice combines crash investigation, exposure analysis, road safety auditing, behavioural science, speed management, conflict observation, vehicle technology, enforcement, and institutional governance.

Road crashes are both a public-health problem and a transport-system problem. Their consequences include fatalities, disabling injuries, emergency-service costs, congestion, productivity losses, psychological trauma, and inequitable impacts on pedestrians, cyclists, motorcyclists, children, older people, and people with disabilities. A technically adequate road may nevertheless be unsafe if it encourages excessive speed, creates ambiguous priorities, provides inadequate crossing opportunities, or fails to accommodate human limitations.

The central professional challenge is therefore to design and manage a forgiving transport system. The system should anticipate mistakes, reduce the probability of conflicts, and prevent ordinary human errors from producing fatal consequences.

Learning Objectives

By the end of this chapter, the learner should be able to:

1. Explain the safe-system approach and its relationship to conventional crash reduction.

2. Analyse crash data using exposure, severity, collision-pattern, and network-screening measures.

3. Apply road safety audit principles throughout the project life cycle.

4. Explain traffic-conflict techniques and their strengths and limitations.

5. Relate human perception, decision-making, workload, and behaviour to road design.

6. Evaluate speed-management and vulnerable-road-user treatments.

7. Select and assess countermeasures using evidence, implementation feasibility, equity, and ethical criteria.

8. Recognise the appropriate and inappropriate uses of digital technologies and artificial intelligence in safety management.

1. Foundations of Road Safety Thinking

1.1 From crash prevention to safe-system thinking

The traditional approach often treats crashes as isolated events caused primarily by driver error. Investigation then focuses on enforcement, education, or correcting a site with a high historical crash count. These measures may be useful, but the approach can imply that the road user alone bears responsibility.

Safe-system thinking begins with four propositions:

• people make mistakes;
• the human body has limited tolerance to crash forces;
• responsibility is shared among designers, operators, regulators, vehicle manufacturers, emergency services, and users; and
• the transport system should be designed so that common errors do not result in death or serious injury.

The safe system does not seek to eliminate all crashes, an unrealistic objective. It seeks to eliminate fatal and life-changing outcomes through safer speeds, safer roads, safer vehicles, safer users, and effective post-crash care. This is closely associated with the Vision Zero principle, although a policy commitment to zero deaths must be supported by measurable actions rather than treated as a slogan.

1.2 The road-safety chain

Safety performance can be considered across a chain:

1. Exposure: how much travel occurs, by whom, where, and at what times.

2. Risk: the probability that exposure produces a crash or conflict.

3. Crash severity: the probability that a crash produces a serious injury or death.

4. Post-crash outcome: the speed and quality of rescue, treatment, and rehabilitation.

A reduction in total crashes is not the only measure of success. A treatment that changes a small number of severe crashes into more frequent but minor property-damage events may improve safety, even if the crash count rises. Conversely, a reduction in reported crashes may reflect under-reporting rather than real improvement.

1.3 Important terminology

A crash is a collision or other harmful event involving a road user, vehicle, object, or roadway environment. The term is preferred to “accident” because most crashes are influenced by identifiable and preventable conditions. A fatality is a death resulting from a crash under the reporting definition used by the relevant jurisdiction. Serious injury requires a clearly defined medical or administrative classification; definitions must not be mixed across datasets.

A collision diagram displays crash locations and movements, while a condition diagram records geometric, operational, roadside, and environmental features. A black spot is a location or short route segment identified through a locally defined high-crash or high-risk criterion. A hazard is a condition with potential to cause harm. A risk combines the likelihood of an undesirable event with its consequences. A conflict is an interaction in which road users must take evasive action to avoid a collision, whether or not a crash occurs.

2. Crash-Data Analysis for Safety Decisions

2.1 Data structure and quality

A useful crash database links at least four categories of information:

• Event data: date, time, weather, lighting, crash type, injury severity, vehicles, and road users.
• Location data: coordinates, route, direction, lane, intersection approach, and nearby land use.
• Exposure data: traffic volume, pedestrian and cyclist counts, vehicle kilometres travelled, public-transport activity, and population or employment served.
• Context data: geometry, speed limits, actual speeds, enforcement, construction, maintenance, and changes in operation.

Data quality problems are common. Police records may omit minor crashes, location coordinates may be inaccurate, injury classifications may differ between hospitals and police, and changes in reporting systems may create artificial trends. Analysts should document missing values, duplicate records, changes in definitions, and the period during which a site had comparable geometry and traffic control.

2.2 Rates and exposure

Raw crash counts are easy to communicate but can be misleading. A location carrying twice as much traffic may experience more crashes simply because it has greater exposure. A simple crash rate per million entering vehicles at an intersection is:

\[

R = \frac{C\times 10^6}{V}

\]

where \(R\) is crashes per million entering vehicles, \(C\) is the number of relevant crashes during the analysis period, and \(V\) is the number of entering vehicles during the same period. If annual traffic is used, both \(C\) and \(V\) must refer to one year, or both must be consistently scaled for a different period.

For a road segment, a commonly used rate per million vehicle-kilometres is:

\[

R_s = \frac{C\times 10^6}{ADT\times 365\times L}

\]

where \(ADT\) is average daily traffic in vehicles per day and \(L\) is segment length in kilometres, assuming the crash count covers one year. If the period is not one year, the denominator must use the actual number of days. These rates support comparison but do not, by themselves, prove that a site is unusually hazardous.

For vulnerable-road-user analysis, exposure may be pedestrian crossings, cyclist trips, or person-hours rather than vehicle volume. A rate based only on vehicle traffic can obscure risk to people walking or cycling.

2.3 Severity and collision patterns

Crash analysis should separate frequency from severity. Useful summaries include the number and proportion of fatal, serious-injury, minor-injury, and property-damage-only crashes; a severity index; and the distribution of collision types. A weighted severity index may be used only when the weights are explicitly defined by the agency. Because arbitrary weights can distort priorities, serious-injury and fatal outcomes should normally be reported separately rather than hidden in one composite score.

Collision patterns often reveal mechanisms that total counts conceal. Examples include:

• rear-end crashes associated with queues, sudden stopping, or inadequate sight distance;
• right-angle crashes associated with crossing conflicts and failure to yield;
• turning and crossing crashes involving pedestrians or cyclists;
• run-off-road crashes associated with speed, curvature, roadside hazards, or loss of control;
• head-on crashes associated with opposing traffic separation and overtaking;
• dooring and side-swipe crashes associated with cycling facilities and parking design.

Temporal analysis should examine hour, day, season, school periods, darkness, rainfall, and special events. Spatial analysis should consider not only the crash point but also the approach, queue, crossing, merging area, or roadside feature that contributed to the event.

2.4 Network screening and regression to the mean

Network screening identifies locations for investigation. Methods include crash frequency, crash rate, severity weighting, empirical Bayes estimates, and models based on expected crash frequency. Frequency methods favour busy sites; rate methods can overemphasise low-volume sites with unstable rates. A mature programme uses more than one measure and then conducts engineering review.

A critical statistical issue is regression to the mean. A site may be selected because it experienced an unusually high number of crashes in one period. Even without treatment, its crash count may decline in the next period as random variation returns toward the long-term average. If the decline is incorrectly attributed to a treatment, the estimated benefit is biased upward. Before-and-after studies should use comparison groups, longer observation periods, or empirical Bayes methods where suitable data exist.

Crash migration is another concern. A treatment may shift crashes to nearby intersections or parallel routes. Evaluation should therefore define a study area wider than the treated point and examine changes in traffic patterns and exposure.

2.5 Illustrative calculation

An intersection records 18 injury crashes in three years. It has an average of 30,000 entering vehicles per day. The three-year exposure is:

\[

V = 30{,}000\times 365\times 3 = 32{,}850{,}000 \text{ entering vehicles}.

\]

The injury-crash rate is therefore:

\[

R = \frac{18\times 10^6}{32{,}850{,}000}=0.548

\]

or approximately 0.55 injury crashes per million entering vehicles. This number is meaningful only when compared with similar intersections and interpreted alongside crash types, severity, traffic control, pedestrian exposure, and site conditions. It is not a universal safety threshold.

3. Road Safety Audits and Proactive Risk Management

A road safety audit is a formal, independent, multidisciplinary examination of a road project or existing road from the perspective of all road users, aimed at identifying crash potential and opportunities to reduce severity. An audit is not a design approval, code-compliance check, liability investigation, or guarantee that crashes will not occur.

3.1 Audit stages

Audits may be conducted at:

• feasibility and planning;
• preliminary or concept design;
• detailed design;
• pre-opening;
• early operation; and
• existing-road or route-network stages.

Early audits are often most influential because alignment, access strategy, cross-section, intersection form, and speed environment can still be changed. A pre-opening audit identifies construction and operational issues such as confusing markings, signs, lighting, barriers, drainage, pedestrian continuity, and unexpected priority arrangements.

3.2 Audit process

A robust process includes:

1. commissioning with a clear scope and available information;

2. assembling an independent, suitably experienced team;

3. reviewing drawings, traffic data, operating-speed information, adjacent land use, and relevant standards;

4. conducting day and night site inspections where appropriate;

5. considering each user group and movement, including people with disabilities;

6. identifying hazards, likely crash mechanisms, and severity implications;

7. preparing findings with evidence, not merely listing non-compliances;

8. obtaining a formal response from the project owner; and

9. verifying implementation or documenting accepted residual risks.

Audit findings should distinguish a hazard from its mechanism and treatment. For example, “pedestrian crossing exposed to high approach speeds” is stronger than “crossing unsafe.” The auditor should explain who is affected, under what conditions, how a conflict may occur, and how the consequence could be reduced.

3.3 Human-factors audit questions

An auditor should ask whether road users can:

• detect and recognise the road environment in time;
• understand priority and permitted movements;
• select a safe path without excessive workload;
• perform the required manoeuvre within available space and time;
• recover from a mistake; and
• see and be seen in darkness, glare, rain, or obstruction.

This approach identifies latent design risks that may not yet appear in crash data.

4. Traffic-Conflict Techniques

Crash data are retrospective and often sparse. Conflict techniques provide a prospective or complementary method by observing interactions that nearly result in collisions. A conflict may be classified by evasive action, time proximity, trajectory, or severity.

4.1 Time-to-collision and post-encroachment time

If two road users are on a collision course and their relative closing speed is approximately constant, a simplified time-to-collision measure is:

\[

TTC = \frac{d}{v_r}

\]

where \(d\) is the relevant separation distance and \(v_r\) is the closing speed, with consistent units. The formula is meaningful only when the projected paths actually intersect and the relative motion assumption is reasonable. It is not appropriate when the road users are diverging, one has stopped, or trajectories are highly uncertain.

Post-encroachment time is the time between one road user leaving a conflict area and another arriving at that area. Low values indicate close temporal interaction, but thresholds vary by conflict type, speed, road-user mode, and observation method. No single numerical threshold should be transferred between jurisdictions without validation.

4.2 Strengths and limitations

Conflict observation can identify emerging risks soon after a geometric or operational change, including sites with few crashes because exposure is low. Video analysis can measure speeds, paths, yielding, crossing gaps, and evasive actions. However, observers may disagree; camera position and weather affect visibility; near-conflicts do not always predict crashes; and a treatment may alter behaviour in ways that reduce observable conflicts without improving severe-crash risk. Conflict studies should therefore use defined protocols, trained coders, reliability checks, and, where possible, crash and speed data.

5. Human Factors in Road Transport

Human factors examines how people perceive, decide, act, and recover within a system. It is not equivalent to blaming users for error.

5.1 Perception and conspicuity

Road users require adequate time and visual information to detect a hazard, interpret its significance, choose a response, and act. Detection is affected by contrast, clutter, lighting, weather, vehicle obstruction, age, fatigue, and expectation. A sign can meet dimensional standards yet be ineffective if it is hidden among advertisements or placed where attention is directed elsewhere.

Conspicuity is not the same as visibility. An object may be visible but not noticed because it is unexpected or competes with other stimuli. Design should support both detection and recognition through clear hierarchy, appropriate contrast, consistent placement, and avoidance of unnecessary visual clutter.

5.2 Expectation, workload, and decision errors

Drivers form expectations from road geometry, markings, previous experience, and surrounding traffic. Inconsistent geometry or an unexpected priority change can produce errors even among attentive drivers. Excessive information increases workload, while insufficient information creates uncertainty. At intersections, simple and legible channelisation is usually safer than a visually complex arrangement requiring rapid interpretation.

Errors may be classified as slips or lapses, in which the intended action is not executed correctly, and mistakes, in which the plan itself is inappropriate. A safe design accommodates both. Examples include forgiving roadside zones, median separation, speed reduction, redundant warnings, and clear recovery space.

5.3 Speed, perception, and injury risk

Speed affects stopping distance, field of view, time available for response, crash energy, and injury severity. Kinetic energy is proportional to the square of speed:

\[

E_k=\frac{1}{2}mv^2

\]

where \(m\) is mass and \(v\) is speed in metres per second. This physical relationship does not by itself predict injury outcome, because crash configuration, occupant protection, vehicle compatibility, and impact surface also matter. Nevertheless, it explains why modest speed reductions can substantially reduce crash energy.

Speed is also a behavioural and environmental outcome. Posted limits alone may not control operating speed when the road is wide, visually open, and free of roadside friction. Effective management aligns the speed limit, road function, geometry, roadside activity, enforcement, and public communication.

6. Speed Management

Speed management has three linked components: setting credible limits, designing for the desired operating speed, and ensuring compliance.

6.1 Setting and communicating limits

A speed limit should reflect road function, roadside access, vulnerable-user activity, intersection density, visibility, crash history, vehicle mix, and the tolerable injury risk of likely conflicts. The operating speed of existing traffic is relevant but should not be the sole determinant; using the prevailing speed to justify a high limit can reinforce an unsafe design.

Limits should be consistent, understandable, and supported by transition treatments. Abrupt changes without visual or physical reinforcement often produce poor compliance. Variable limits may be useful where conditions change substantially, but they require reliable detection, clear signs, enforcement, and public comprehension.

6.2 Self-enforcing design

Speed humps, raised intersections, narrow lane arrangements, gateway treatments, curvature, median refuge design, tighter corner radii, and streetscape friction can influence speed. Each treatment must be checked for emergency access, buses, cyclists, drainage, noise, accessibility, and effects on adjacent roads. A speed treatment that displaces high-speed traffic into a residential side street is not a complete solution.

7. Protecting Vulnerable Road Users

7.1 Pedestrians

Pedestrian safety requires continuous, accessible routes; safe crossing opportunities; adequate sight distance; manageable crossing distance; appropriate lighting; and speeds compatible with human injury tolerance. Treatments may include raised crossings, refuge islands, curb extensions, signal control, pedestrian phases, access management, improved lighting, and removal of parking near crossings.

A crossing should be located where people need to cross, not solely where it is convenient to install. If desire lines are ignored, pedestrians may cross outside the facility. Designers should account for children, older people, wheelchair users, people with visual impairments, and people carrying goods.

7.2 Cyclists and motorcyclists

Cyclist safety depends on network continuity, intersection treatment, separation where speed differentials are high, predictable priority, surface quality, drainage, and protection from opening vehicle doors. Facilities should avoid abrupt termination and should make turning movements legible.

Motorcyclists are vulnerable to loss of control, poor surface conditions, roadside obstacles, unsafe barriers, and visibility problems. Maintenance, skid resistance, safe barrier design, drainage, and treatment of utility covers can be as important as geometric redesign.

7.3 Public transport users and freight interactions

Bus stops should provide safe boarding areas, pedestrian access, adequate visibility, and conflict-minimising relationships with cycle tracks and general traffic. Freight loading, deliveries, and heavy-vehicle turning require attention to blind zones, swept paths, reversing, and pedestrian separation. Safety assessments should include the whole trip to and from a stop or destination, not only the vehicle movement.

8. Selecting and Evaluating Countermeasures

Countermeasure selection should be based on a diagnosed crash mechanism, not a catalogue of fashionable treatments. A useful decision sequence is:

1. define the safety problem and affected users;

2. verify exposure, crash patterns, speed, and site conditions;

3. identify the underlying mechanism;

4. generate several alternatives, including systemic and low-cost options;

5. assess expected safety benefit, severity reduction, accessibility, equity, cost, constructability, maintenance, and network effects;

6. select a treatment or package;

7. implement with a monitoring plan; and

8. evaluate outcomes and revise if necessary.

8.1 Before-and-after evaluation

A simple percentage change is:

\[

\%\Delta = \frac{Y_{after}-Y_{before}}{Y_{before}}\times100

\]

where \(Y\) is the chosen outcome, such as severe crashes, provided the periods and exposure are comparable. This calculation is descriptive, not automatically causal. It can be distorted by regression to the mean, traffic-volume changes, weather, enforcement, economic conditions, reporting changes, and unrelated projects.

Stronger evaluations use comparison sites, exposure adjustment, interrupted time-series methods, or empirical Bayes approaches. Outcomes should include severity, user-specific effects, speeds, yielding, accessibility, and unintended consequences. A treatment that reduces total crashes but increases serious cyclist injuries requires reconsideration, even if its aggregate statistic looks favourable.

8.2 Economic and equity appraisal

Benefit–cost analysis can support prioritisation, but monetary values should not be interpreted as the value of a person’s life in a moral sense. They are administrative values used to compare public investments. Distributional analysis is essential: a project may have a favourable aggregate benefit while imposing risk or access burdens on low-income communities, people with disabilities, or informal transport users.

9. Illustrative Case: Safety Transformation of a Suburban Arterial

The following is a fictional illustrative case based on common professional conditions.

Background

A four-lane suburban arterial served a secondary school, shops, and two bus routes. Over five years, the corridor recorded 42 reported crashes, including two fatal pedestrian crashes, six serious-injury crashes, and numerous rear-end and turning crashes. The posted speed limit was 60 km/h, and observed 85th-percentile speed was 67 km/h. Pedestrians crossed between bus stops and shops, although the nearest signalised crossing was 280 m away.

Problem analysis

Crash mapping showed that the fatal and serious pedestrian crashes occurred near informal crossing points. Video observations recorded high approach speeds, gaps that were difficult for children and older people to use, and vehicles stopping in the curb lane near the shops. A safety audit identified poor night lighting, wide turning radii, no pedestrian refuge, inconsistent bus-stop design, and visual clutter. The analysis concluded that the problem was systemic: high speed, excessive crossing distance, inadequate crossing location, and competing roadside activity interacted to create severe conflicts.

Approach

The project team developed a package rather than a single treatment:

• a raised, signal-controlled pedestrian crossing aligned with the principal desire line;
• a central refuge and reduced crossing distance;
• narrower lanes and tighter corner geometry at side roads;
• relocation and formalisation of bus stops;
• improved lighting and removal of parking near the crossing;
• a reduced speed limit supported by gateway markings and enforcement;
• continuous accessible footpaths and protected cycle continuity; and
• a twelve-month monitoring programme covering speeds, yielding, conflicts, crashes, bus operations, and access for people with disabilities.

Outcome

After one year, mean approach speed fell by 8 km/h, yielding increased substantially, and observed serious conflicts declined. No fatal or serious-injury pedestrian crash occurred during the first year, although the observation period was too short to establish a definitive long-term crash effect. Rear-end minor crashes increased slightly during the initial adjustment period. The agency retained the treatment, refined signal timing, and continued evaluation rather than declaring permanent success from one year of data.

Learning points

The case demonstrates that crash data alone did not identify the full mechanism; video, audit, speed, and accessibility evidence were necessary. It also shows why speed management, crossing location, public transport, geometry, and enforcement should be designed as a package. Finally, evaluation must consider both safety benefit and transitional or unintended effects.

10. Technology, Data Governance, and Emerging Practice

Automated video analytics, connected vehicles, smartphone data, vehicle event recorders, geographic information systems, and digital twins can improve safety diagnosis. Computer vision may estimate trajectories, speeds, near-conflicts, queue lengths, and pedestrian volumes at a scale difficult to achieve manually. Machine-learning models can screen networks for patterns associated with risk.

Technology does not remove the need for engineering judgement. Algorithms may perform poorly in rain, darkness, unusual road layouts, or locations under-represented in training data. A model trained on vehicle crashes may systematically overlook pedestrian fear, accessibility barriers, or near-misses involving children. Agencies should validate accuracy, report uncertainty, maintain human review, and test for unequal performance across road-user groups.

Privacy and governance are central. Video and connected-device data should be collected for a legitimate purpose, minimised, secured, de-identified where possible, retained for defined periods, and governed by transparent access rules. Automated enforcement must provide due process and reliable evidence. Safety innovation should not become a justification for indiscriminate surveillance.

11. Challenges and Professional Responsibilities

Several challenges complicate safety practice:

• crash data are incomplete and delayed;
• rare severe crashes create statistical uncertainty;
• safety treatments may transfer risk geographically or between road-user groups;
• political pressure may favour visible projects over effective ones;
• standards represent minimum requirements, not proof of optimal safety;
• changes in mobility, climate, land use, and vehicle technology alter exposure and risk; and
• agencies often have fragmented responsibility for roads, policing, health, vehicles, and emergency response.

Professionals should communicate uncertainty honestly, avoid presenting correlation as causation, preserve an audit trail for decisions, and disclose conflicts of interest. Ethical practice requires giving appropriate weight to people who have less protection in a collision and less influence in infrastructure decisions. Safety must be integrated with accessibility, environmental quality, public health, and social inclusion.

Best-Practice Framework

A defensible safety programme should:

1. adopt a safe-system objective focused on preventing fatal and serious injury;

2. combine crash, exposure, speed, conflict, design, and community evidence;

3. audit projects early and independently;

4. design for human limitations and recovery from error;

5. manage speed through credible limits, self-enforcing design, and enforcement;

6. protect vulnerable users through continuous, accessible networks;

7. evaluate countermeasures with appropriate comparison and exposure data;

8. examine equity, privacy, maintenance, and unintended consequences; and

9. treat implementation as an iterative learning process rather than a one-time project.

Chapter Summary

Transportation safety is a system property produced by road users, vehicles, infrastructure, speeds, institutions, and post-crash care. Safe-system thinking recognises human fallibility and seeks to prevent ordinary errors from causing fatal or serious injury. Crash analysis should combine frequency, severity, collision patterns, exposure, and context while accounting for data quality, regression to the mean, and crash migration.

Road safety audits provide proactive, multidisciplinary review from planning through operation. Conflict techniques complement crash records by examining evasive interactions, but their measures require defined assumptions and validation. Human-factors analysis addresses perception, conspicuity, expectation, workload, speed, and recovery from error. Speed management requires coordination among limits, design, enforcement, and land-use activity. Vulnerable-user protection depends on continuous accessible networks, appropriate crossing facilities, safe intersections, visibility, and tolerable impact speeds.

Countermeasures should be selected from diagnosed mechanisms and evaluated using rigorous before-and-after methods, comparison evidence, exposure measures, and equity analysis. Digital technologies can improve observation and prediction, but they require validation, privacy protection, transparency, and human oversight. The professional goal is not merely fewer reported crashes; it is a fair, accessible, resilient transport system in which fatal and life-changing outcomes are prevented.

Key Terms

Safe system: A transport approach that accepts human error but seeks to prevent fatal and serious-injury outcomes.

Vision Zero: A policy principle aiming to eliminate deaths and serious injuries in the transport system.

Exposure: The amount and type of travel opportunity or activity subject to risk.

Crash rate: Crash frequency normalised by a measure of exposure.

Severity: The level of harm resulting from a crash.

Collision diagram: A graphical representation of crash locations and movements.

Black spot: A location or route segment identified as having unusually high crash occurrence or risk.

Regression to the mean: The tendency for unusually high or low observations to move toward their longer-term average.

Road safety audit: An independent, systematic examination of a road project or road in operation to identify safety opportunities.

Hazard: A condition capable of causing harm.

Conflict: A road-user interaction requiring evasive action to avoid a collision.

Time-to-collision: Estimated time until collision if projected movement and closing speed remain applicable.

Post-encroachment time: Time between one road user leaving a conflict area and another arriving.

Human factors: Study of human capabilities, limitations, behaviour, and interaction with systems.

Conspicuity: The extent to which an object or road feature attracts attention and is noticed.

Operating speed: The speed at which vehicles actually travel under specified conditions.

Vulnerable road user: A road user with relatively high exposure to injury, including pedestrians, cyclists, motorcyclists, children, and older people.

Countermeasure: An intervention intended to reduce crash probability or severity.

Empirical Bayes evaluation: A statistical approach that combines observed crash data with expected values to reduce estimation bias.

Exposure adjustment: Modification of a safety measure to account for changes in the amount of travel or activity.

Review Questions

Short Answer Questions

1. What are the principal components of the safe-system approach?

2. Why can raw crash counts be misleading when comparing sites?

3. Define regression to the mean and explain its relevance to safety evaluation.

4. What is the purpose of a road safety audit?

5. Distinguish time-to-collision from post-encroachment time.

Long Answer Questions

1. Explain how human perception, expectation, workload, and error influence road design and operation.

2. Discuss the stages, team requirements, and outputs of a road safety audit.

3. Explain why speed management requires more than changing a posted speed limit.

4. Compare retrospective crash analysis with prospective conflict analysis, including limitations of each.

5. Describe an evidence-based framework for selecting and evaluating road safety countermeasures.

Analytical / Application Questions

1. A corridor has many minor rear-end crashes but few serious crashes, while a nearby low-volume intersection has three fatal crashes in four years. Propose a screening and treatment strategy that avoids relying on crash frequency alone.

2. Design a safety-audit approach for a proposed urban arterial crossing a school and a bus corridor. Identify the human-factors, accessibility, speed, and governance issues that should be examined.

3. A before-and-after study reports a 30% crash reduction one year after treatment, but traffic volume fell by 20% and a nearby parallel road received a new signal. Explain how you would investigate whether the treatment caused the apparent improvement.

Multiple Choice Questions with Answers

Question 1

Which principle is central to safe-system thinking?

A. Road users are solely responsible for preventing crashes

B. All crashes can be eliminated through driver education alone

C. The transport system should reduce the consequences of human error

D. Crash frequency is the only valid safety measure

Correct Answer: C. The transport system should reduce the consequences of human error

Question 2

An intersection has 18 injury crashes over three years and 30,000 entering vehicles per day. What is the approximate injury-crash rate per million entering vehicles?

A. 0.055

B. 0.55

C. 5.48

D. 54.8

Correct Answer: B. 0.55

Question 3

What is the main purpose of exposure adjustment in crash analysis?

A. To remove all severe crashes from the dataset

B. To account for differences in the amount of travel or activity

C. To replace field inspections

D. To guarantee that a treatment is effective

Correct Answer: B. To account for differences in the amount of travel or activity

Question 4

Regression to the mean can cause analysts to:

A. Underestimate all future crash counts

B. Attribute a natural fluctuation in crashes to a treatment

C. Ignore traffic-volume changes entirely

D. Treat conflicts as identical to crashes

Correct Answer: B. Attribute a natural fluctuation in crashes to a treatment

Question 5

A road safety audit is best described as:

A. A legal investigation after a fatal crash

B. A formal independent examination aimed at identifying safety opportunities

C. A replacement for detailed engineering design

D. A guarantee that no crashes will occur

Correct Answer: B. A formal independent examination aimed at identifying safety opportunities

Question 6

The simplified time-to-collision equation d divided by relative closing speed is applicable when:

A. The road users are diverging

B. Their projected paths intersect and relative motion is reasonably constant

C. One road user is permanently stationary outside the path

D. Only annual crash counts are available

Correct Answer: B. Their projected paths intersect and relative motion is reasonably constant

Question 7

Which intervention most directly addresses pedestrian risk created by high crossing speeds and long crossing distances?

A. Increasing roadside advertising

B. Providing a refuge or raised crossing with speed reduction

C. Removing all pedestrian signs

D. Increasing the lane width

Correct Answer: B. Providing a refuge or raised crossing with speed reduction

Question 8

Why is a posted speed-limit change alone often insufficient?

A. Operating speed is influenced by road design, environment, and enforcement

B. Speed limits never affect driver behaviour

C. Crash severity is unrelated to speed

D. Speed limits apply only to heavy vehicles

Correct Answer: A. Operating speed is influenced by road design, environment, and enforcement

Question 9

Which statement about conflict techniques is most accurate?

A. They always predict crashes more accurately than crash records

B. They can reveal emerging interactions but require defined protocols and validation

C. They eliminate the need for exposure data

D. They are valid only where fatalities have occurred

Correct Answer: B. They can reveal emerging interactions but require defined protocols and validation

Question 10

A responsible use of AI-based video safety analytics requires:

A. Deployment without validation because algorithms are objective

B. Collection of identifiable data indefinitely

C. Accuracy testing, uncertainty reporting, privacy safeguards, and human oversight

D. Replacing all professional judgement with automated rankings

Correct Answer: C. Accuracy testing, uncertainty reporting, privacy safeguards, and human oversight